- 1 The research for this article was made possible by a grant from the Central Research Fund of the Un (...)
1The present paper originally appeared in the context of the panel named «Political Structures» of the AEGIS Conference. However, the materials upon which it is based come from the fieldwork undertaken for my doctoral thesis, which was carried out for the most part in a small town in Northern Mali1. My main concern was thus with politics as it is enacted in a local setting, or in other words, with political action. On the other hand, the fieldwork took place right after the setting up of the new municipal structures created by the reform of decentralisation in Mali, and this reform did of course modify the conditions for the exercise of local politics.
2These modifications were not necessarily of the kind promoted by the reform, however. In fact, there are important discrepancies between the official version of the decentralisation on the one hand and its enactment and the ways in which it is being perceived on the other. This is why I shall largely, but not completely, leave the changes in the official politico-administrative structures in Mali to one side. In my opinion, an assessment of the reform in question that overestimates the importance of such official structures will most likely overlook other, potentially more significant, developments.
3This is not to say that I intend to privilege the local for the sake of the local, or that I think any given locality is somehow insulated from processes external to it. Rather, I shall argue that through close attention to developments on the local level, one may become aware of rather more insidious changes, not least to the ways in which the state is operating. Behind the reform of decentralisation, then, state institutions take on new and for the most part reduced roles, thereby opening up space for new forms of political organisation. From this point of view, one may even say that a new structure is gradually appearing. This will also constitute my main point of interest in this paper.
4For such an argument to be intelligible, however, a certain amount of background information will be required. Below, I shall therefore first briefly discuss the reform of decentralisation in its official form as well as explain why it is being undertaken. There are in Mali quite particular reasons for the reform, linked to a rebellion that took place in Northern Mali. The feature that both the decentralisation and the rebellion have in common is a concern with and wish for more political influence to disfavoured constituencies. A concern with access to state resources is also of importance, however, although the significance of such resources changes as the rebellion gives way to the reform. Having provided this background information, I shall then proceed to a discussion of local politics in the town of Gossi. Through the use of a few examples, the new relationship between the state and the localities will become apparent. Anticipating my conclusion a little, I may say that the state increasingly concentrates its activities on the realm of control and oversight, while reducing its role as a provider of resources. Ambitious actors are thus free to take over the local arena, while relying more and more on funds deriving from sources other than the state. For the majority of the people, devoid of this requisite organisational capital, «participation», which is allegedly made possible by the decentralisation, consists henceforth primarily in building a relationship with one such influential figure. The result is an increasing fragmentation where each group of people backing such a leader tends to become a semi-autonomous community.
5Decentralisation has become something of a trend in many parts of the world. In Africa, the notion appears to have gained increasing favour following the growing emphasis on democratisation from the early 1990’s onwards, and thus to have become part of a conceptual package including democratisation and good governance as well as participation. The latter form part of an international discourse prevalent among African governments and among donors such as the World Bank and the IMF (e.g., Manor 1999).
6Decentralisation as a concept nevertheless denotes very different projects of reform. In the literature on the subject, despite debates over the apposite nomenclature, there is a consensus that a process of decentralisation should entail more than a simple deconcentration of state services; it must involve a real devolution of powers of decision-making (e.g., Mahwood 1993, Manor 1999). The question still remains to what level power will be devolved and to what organ. Only a decentralisation to popularly elected councils may be equated with democratisation in the general acceptance of the latter.
7Compared to other African countries engaged in processes of decentralisation, Mali has chosen a radical approach to the issue. The reform process involves the creation of municipal councils elected by general suffrage. The municipalities correspond to what was formerly the lowest echelon of the state administration, and their modest size should ensure the accessibility of the local authorities. The former system, meanwhile, placed all decision-making powers locally in the hands of centrally appointed state administrators who were only accountable to their superiors (Républic du Mali 1999a).
- 2 Jeune Afrique did for a long time refer to the «Malian model» before adopting a more critical view (...)
8The stated purpose of the reform, then, is to remedy the lack of popular influence on the management of local affairs and the distance separating the authorities and the people (République du Mali 1999b). These being the intentions, the people responsible for the elaboration of the reform commonly present it as the crowning achievement of the country’s process of democratisation (Sall 1993). Ever since the overthrow of the single party regime of Moussa Traoré in 1991 and the subsequent exemplary transition to multi-party democracy, the Malian government has been anxious to promote the democratic credentials of the country. This image has been propagated in the media2, but also through close co-operation with the UNDP and the Bretton Woods institutions (see Poulton & Youssouf 1998). Mali has thus become a display case for democracy in West Africa. This in turn ensures continued funding from the major donors and from NGO’s, as the country now ostensibly conforms to the new conditionalities for external funding, which rest on successful democratisation.
- 3 A similar argument is made by Seely (2001).
- 4 The following account of the rebellion is based on Boilley 1999, Diarrah 1996 and Poulton and Ag Yo (...)
9To democratisation, whether in earnest or as a showcase, must be added one more factor, in the case of Mali This is the rebellion, which, although it did probably not influence the decision to decentralise, nevertheless affected the scope and the timing of the reform process3. The rebellion was of too complex a nature for me to discuss it in much detail here. I shall only give a brief summary of its causes and development, before bringing to the fore the issues of political influence and access to resources that matter most for my argument. The rebellion started when, in 1990, exiled Tuareg based in Libya and Algeria revolted against the central state by attacking army posts4. Through such attacks they maintained their supplies and armaments, but were never able to establish much territorial control. Repression against the civilian Tuareg and Arabs ensued, as the army was unable to hunt down the rebels, and a refugee crisis was the result. Two peace agreements were negotiated between succeeding governments and the rebels, but only partly implemented. In the midst of growing ethnic polarisation, some rebel leaders managed to approach notables of other ethnic groups and reach a new deal with the government. The end result was an offer of positions in the state, and particularly in the army, to the rebels and not least to their leaders, after which the conflict died down while the refugees gradually returned to the country.
10The rebels had all along made claims of an independent state for the Tuareg and the Arabs, a state carved out from the territory of the three northern regions of Mali. Through the course of the rebellion, some rebels tempered their demands to one of autonomy, provisions for which were included in the second peace agreement, which remained the final one, albeit with modifications. The northern regions had thus been granted a so-called special status. Intense opposition from other regions of the country was the result, however, and the notion of a special status may even have fuelled the conflict. When it ended in 1996, therefore, the decision was made to implement the process of decentralisation as soon as possible in all regions of the country. This was justified on the basis that the decentralisation allowed a large measure of local autonomy. The people of Northern Mali who had revolted against exaggerated state interference could now freely manage their own affairs, it was held.
11The grievances of the rebels involved certain other issues, however. In spite of the calls for autonomy, they consistently laid claim to a larger proportion of state development funds and positions in the state, on the basis of alleged discrimination. And the rebellion did finally end by their being granted a greater share of state resources.
12There is thus an apparent inconsistency between claims of independence and of stronger incorporation into the Malian state. It may be attributed to differences of opinion among the rebels and changing strategies throughout the course of the conflict. On the other hand, such apparently contradictory demands may reflect a wish for a state capable of facilitating a process of development rather than a state simply exercising control. Unfortunately, the state was already characterised by the weight of its security apparatus, and the rebels reinforced this imbalance. Already hard pressed for resources even without the rebellion, the state thus concentrates increasingly on the task of surveillance. The one of development, meanwhile, is turned over to the decentralised entities and their external partners, as the following case study will show.
13My fieldwork location, Gossi, is a small town in the region of Timbuktu. Since the droughts of the 1980’s, it has been the host of several NGOs because of its central location next to the only paved road traversing the area. This location and the NGO presence have both contributed to the growth of the town and have turned it into a regional centre. Its livestock market attracts pastoralists from all the neighbouring localities and is of central importance in an area where the economy is to a large extent based on pastoralism. A trading and administrative centre, the town has attracted people from other regions of the country as the NGOs drew drought victims here. The result is considerable complexity in the ethnic composition of the local population and its social organisation.
14This ethnic diversity is of significance in local politics, of course, but not quite in the way one would expect. Local political struggles are not simply factional ones opposing the main ethnic groups or first-comers and latecomers. In fact, the most visible opposition in ethnic guise matters but little in local politics. This is the opposition between the people most closely associated with the rebellion, the Tuareg and the Arabs who are known as the «Red», and the remaining ethnic groups the members of which are known as the «Black». The rebellion did to some degree exacerbate this opposition, but earlier relationships of co-operation across such boundaries have persisted despite the conflict.
15This is not to say that local communities based on ethnic groups or subgroups are of no importance. For a host of activities, such communities are the obvious frame of reference. In the local discourse, therefore, local politics is often presented as a struggle between local groups, while an important aim is to be represented in the main decision-making organs, most notably the municipal council. However, certain features of local social organisation should make us wary of taking this ethnicised discourse at face value.
16The great mobility of the people coupled with changes to the economic situation since the droughts has already modified many aspects of earlier social organisation. Many sub-ethnic groups now appear in local branches, as it were, that only have relevance in a given locality. Such groups, known as «fractions», alongside their leaders or «chiefs», should normally be recognised by the state administration as being resident in a particular municipality. Both chiefs and fractions, then, in spite of their «traditional» flavour, are technically extensions of the state administration. Chiefs should thus be elected by the fraction members, but they more frequently inherit the position, when they not simply manage to gather a sufficient following to create a new fraction of which they themselves will of course be the chief. Furthermore, the «fractions» quite frequently fission or give rise to sub-divisions, one of the main reasons being struggles over the chieftaincy. Some enterprising chiefs will even include people from socially sub-ordinate categories, such as the descendants of slaves, into their fraction to boost its numbers, which, among other things, adds to its voting strength. There is thus no one single principle such as kinship underlying local social organisation, although local discourse privileges such connections. The resulting complexity entails that there is considerable scope for alliances cross-cutting the most strongly publicised rivalries, as it were.
17The game of alliances engaged in by the chiefs when establishing their positions adds to the complexity of local alliances and rivalries, however. A closer look at the municipal council, to which fraction chiefs supply most of the councillors, should make this clear. Among the councillors, for instance, only three councillors do not belong to the ruling party, ADEMA. All political parties habitually channel resources towards their local branches to build support, but the ruling party has, unsurprisingly, more to offer than its rivals. Usually, the local members of the party share these resources, although the distribution is affected by the balance of power between them. Two of the councillors representing different parties thus left the ruling party not long before the elections because they had not profited much from party resources.
18In other words, the chiefs in their capacity as councillors co-operate across the boundaries of the fractions they ostensibly represent, and not necessarily to the benefit of their constituencies. In fact, in addition to resources from political parties, they sometimes manage to access those of NGOs by emphasising their role as representatives of their communities. When a wildlife project offered indemnities to victims of damages made by a herd of elephants, the councillors and some associated chiefs swiftly set themselves up as the leaders of the victims, without having suffered any damage themselves. As the example shows, the co-operation of the chiefs in this regard made them benefit from what should have accrued to their fraction members.
19Much the same can be said in the case of the resources of the municipal council itself, even if they cannot be controlled as easily. This is because the state administration still keeps its representative in every municipality, the person formerly representing the only formally recognised decision-maker locally. His role has been redefined to one of assistance, but it is also his duty to report any irregularities to his superiors, who may in theory dissolve the council in cases of serious mismanagement. In Gossi, the state representative has remained on good terms with the councillors, however. In the past, he relied to a considerable extent on the main fraction chiefs, as the resources at his disposal were too limited for an integral control of the locality. Today, it is thus with his collusion or tacit support that the diversion of council funds is possible.
20In this way, the chiefs appropriate much of the public resources. On the other hand, many people are accustomed to the chiefs’ taking care of communal matters, and within certain limits, it is accepted that the chiefs eat, as it were. This may have been some of the reason why many local people showed little interest in the decentralisation as such, although most had clearly heard about the reform. Their main interest was thus in the personnel of the municipal council. Not everyone was so lenient, however, and there were quite a few voices denouncing the corrupt practices of the chiefs.
21In view of the inactivity of the council, this is scarcely surprising. The council only convened almost a year after the municipal elections. Council staff was either not hired or not paid; only when it came to tax collection did the council exhibit any degree of expediency. The result was that the pressure of taxation increased because of the creation of a series of new taxes, while there was no marked improvement in the services provided by local authorities.
22The most common view was thus that the reform would not change much. No one had asked for it; the reform was simply an external imposition decreed by a foreign power such as France, or, according to a more sophisticated version, the donors. In any case, everything was in the hands of the ruling party which was on the road to becoming a new single party. As the municipality was effectively in the hands of the same constellation of chiefs grouped around the state representative as before, one may agree that little indeed had changed.
23In this situation, the reason that there was not a massive popular protest against the chiefs in spite of there being some criticism of their undertakings is because their support is built on redistribution. Lest I turn this into another sordid tale of corruption, however, I wish to draw your attention to some features of the local economy. A strong dependence on external inputs accounts to some extent at least both for the limited success of the decentralisation and the ongoing fragmentation characteristic of local social organisation.
24Since the droughts of the 1970’s, there has been a growing need for external inputs into the pastoralist economy. Such inputs, typically, come from labour migration, to Côte d’Ivoire for instance, and trade. With an increasing amount of capital in the hands of some traders and politicians, the latter have also invested in livestock. Add to this a growing long-distance trade in livestock, and it becomes clear why the prices of animals have risen making the pastoralists ever more dependent on cash.
25The situation in Gossi illustrates this development admirably. Most of the houses in town are owned by traders from elsewhere, while the autochthonous groups who have largely stuck to livestock raising are losing out even in their own domain as town-based traders acquire herds of their own that are entrusted to hired herders. The success of the traders is based on their access to cash in an economy becoming ever more monetarised. Trade in Gossi is nevertheless subject to the fluctuations of the livestock business, which remains the driving force of the local economy. Only two other sources of cash and credit exist, the state and development agencies. This is as true on the individual level as on the level of the municipality.
26In the case of local authorities, it is clear that their increased autonomy has been accompanied by an expectation of greater self-reliance. In the official rhetoric, this is called «responsibilisation» (République du Mali 1999b). Popular participation entails less state intervention; this, however, applies not just to control, but to the support of development efforts as well.
27It is true that the decentralisation laws state that every transfer of tasks should be accompanied by a concomitant transfer of resources (République du Mali 1999a). The actual transfers, however, remained modest and often arrived late. The sums intended for the setting up of the municipal councils, for instance, only arrived more than a year after the elections.
28The lethargy of the municipal council in Gossi did thus to some extent find its origin in the one of the state administration, which was responsible for the preparations of the reform. In spite of the rather grandiose statements regarding the importance of the decentralisation, then, the state administration did little in the way of practical measures to prepare for the reform. For instance, little information was provided to people outside the urban areas; in fact, the main contribution of the administration was to arrange the elections. Afterwards, little assistance was awarded in the setting up of the newly elected councils, and no provisions had been made for the training of their staff. On the other hand, some training of councillors as well as more general information campaigns had been undertaken by NGOs commissioned by the state administration.
29When the decentralisation is approached in this manner, it appears above all to have served the purpose of reducing state expenditure. This is because the state transfers costly tasks such as health and schooling to the municipalities. Meanwhile it retains all the most valuable categories of land whose natural resources, such as firewood, would constitute important sources of income if handed over to the municipalities. Legal provisions for such transfers that would strengthen the self-reliance of the municipalities have been made, but they are seldom implemented (Rép du Mali 1999a). The only substantial transfer of resources is thus certain categories of taxes. However, these are mainly local taxes that the state administration had difficulties collecting in the past.
30The decentralisation opens another opportunity to the municipalities, however. This is the right to negotiate partnerships with external agencies, independently of the state administration (République du Mali 1999a and 1999b). The dearth of resources locally must now be made up for by inputs from such sources. However, the mode of operation of these agencies tends to fuel the competition between the local groups or fractions and their chiefs.
31Admittedly, many NGOs still work in close association with local authorities and make their plans with the municipal council. This is especially true of the larger NGOs whose interventions last for several years. But there is also a growing number of smaller NGOs whose work is characterised by more punctual interventions and which sometimes decide to start work in the locality on the request of a particular chief. The co-ordination between these NGOs is on the whole insufficient. The growing importance of assistance from various Arab Islamic organisations that combine development interventions and the construction of mosques hardly offers a remedy to this state of affairs.
32The chiefs thus experience the push of the municipal council and its NGO collaborators and the pull of the NGOs that target particular fractions. While in the first context they will tend to make attempts to appropriate or divert the available resources, in the second, the NGO contribution may suffice in itself. In both contexts, of course, the presence of NGOs engenders spin-off effects and employment opportunities, which can make them attractive even when their activities are otherwise of little interest. Put differently, the most significant difference is the following one: Diverted resources must at times be redistributed to ensure continued political support, as when votes are bought with gifts.
33Development efforts targeting an individual fraction or community, on the other hand, foster such support in themselves.
34This raises the question of why the fractional logic, as it were, is to this point self-perpetuating. Historical reasons may be invoked, as well as the ostensibly kinship-based idiom on which the fractions appear to be based. However, as mentioned above, even if local discourse emphasises such rules, practices at variance with them abound. What was earlier the main local groups steadily sub-divide, while some people associate because of long-standing ties of friendship and a history of co-operation and association rather than simply kinship or fraction links. Moreover, what appears as an even more ephemeral kind of group is emerging, the site. Its raison d’être gives us a clue to the mechanism behind the ongoing fragmentation.
35The emergence of sites probably harks back to before the rebellion, but their numbers grew as a result of the conflict. During the repatriation of the refugees, many people were settled in so-called sites or small hamlets to facilitate the provisions of aid. A well, for instance, was of more use where there was a certain concentration of people. Increasingly, people tend to settle in such sites and become all but sedentary as the pastoralist transhumance becomes ever more seasonal and only involves the people directly occupied with the livestock.
36A main concern for the site is to attract an NGO to develop it with wells, a school and a clinic. This is the task of the chief of the site, who may be a fraction chief, but who is often simply a person capable of attracting development funds. Depending on his success in this regard, his wealth and his personality, he may attract more people to his site and potentially more funds as well.
37One could possibly argue that the fragmentation is due to historical causes or local cultural idioms such as the importance of chiefs. But the lack of co-ordination of development efforts and the scarcity of resources locally undoubtedly contribute to the growing prestige of the rather few persons who are able to access the external resources. Furthermore, many external agencies readily accept the claims of chiefs to represent a local group, and treat them as privileged interlocutors.
38There is in any case an ongoing fragmentation, which is all the more serious as the groups in question may also be in competition for access to natural resources such as water and pasture. Disputes between these groups and occasional violent clashes, albeit highly localised ones, have therefore occurred. For the time being, however, there is a number of factors that keeps such a development from reaching its logical conclusion, namely a high number of small communities permanently pitted against each other.
39Firstly, there is the municipal council, and the fact that many external development agencies work in close co-operation with the local authorities. To have access to these resources, a degree of co-operation between the main chiefs is necessary. Secondly, there is the system of control and oversight of the state. The main components locally are the state representative and a few divisions of the armed forces.
40As I have already shown, the state reduces its development efforts to reduce state expenditure, hence the decentralisation, and concentrates instead on control. Ironically, the rebellion has contributed to this development. The demand of the rebels for a more equitable distribution of state resources was granted primarily in the form of positions in the armed forces, which are, I need hardly say, the most profitable positions in the Malian state. As a result the number of the members of the armed forces has grown and consumes a great part of state resources. Meanwhile, so-called residual banditry following the rebellion justified a continued emphasis on the maintenance of law and order (see Poulton & Youssouf 1998).
41Unfortunately, despite the considerable amounts spent on the armed forces, insecurity is widespread throughout the northern regions. In Gossi, the town itself was quite safe because of the presence of a company of guards, while in the bush, the pastoralists victims of thefts of livestock were increasingly left to their own devices. The result is a growing demand for firearms, the numbers of which in circulation locally increased because of the rebellion. Furthermore, the members of the armed forces are compromised by their involvement in criminal activities ranging from extortion of money to collusion with smugglers.
42The ability of the armed forces to counteract the process of fragmentation is thus tenuous at best. A worst case scenario, meanwhile, was suggested to me during fieldwork by the following incident. In Mali, there is a «Day of the Armed Forces». On this day, in the town square in Gossi, the guards demonstrated their close combat skills and their ability to defend the local administrative structures which contain the armed forces arsenals. Similar demonstrations took place throughout the country. They thus made clear that the military arsenals were protected against attacks such as took place during the rebellion and which constituted the most important way of acquiring weapons for the rebels.
43To me, however, the need for the soldiers to defend their own weapons became symbolic of the potential of the armed forces to fuel the process of fragmentation rather than contain it.
44A development such as the one suggested in my last example is perhaps too much of an extrapolation. My purpose in invoking it, however, was to bring attention to a situation characterised by a state that has largely withdrawn from most activities other than control, but with little success even in the latter because the armed forces remain the only profitable way of access to the state. Even the state, however, has steadily less to offer, hence the inefficiency and criminal propensities of the members of the armed forces. I shall end this paper with another example which, although it is in some regards an extreme case, serves to highlight the redistribution of roles among the main providers of resources and the emerging principles of the new social organisation.
- 5 For the solution of this particular conflict, see Jeune Afrique 2097 (2001), pp. 30-32.
45Shortly before my leaving the field in December 2000, a small-scale rebellion broke out in the northernmost region of Mali. Many people considered it an upsurge of the last rebellion and grew afraid that it would spread and become a full-scale one. After a few skirmishes with the army, in which some soldiers were killed, the rebel leader agreed to enter into negotiations with the Malian government5. An independent state was out of the question this time, of course, but after a while the rebel leader settled for his own municipality. He also gave up on the claim of positions in the state. The president was relieved, as the rebels were responsible for the deaths of Malian soldiers. Instead, their leader claimed, and was granted, development funds and the right to negotiate deals with development agencies without state interference.
46Rebellion against the state was initially also a struggle for access to the state. In its last manifestation, rebellion against the state is not so much about autonomy, as it is an attempt to exercise pressure on the state to provide resources, or should I say, give access to them, as these resources do not belong directly to the state. The municipality, meanwhile, may in extreme cases, such as the present one, be completely subordinated to local dynamics, and represent little more than a formal recognition from the state.
47Although a form of local autonomy in which one is left to ones own devices may allow for a degree of local development and economic activity, this is only at the cost of insecurity and political instability resulting from the fierce competition for resources. Long-term development efforts, on the other hand, will be difficult as long as the decentralisation is employed as a way of reducing the responsibility of the state, especially in remote parts of the country. Only a coordination of the development efforts can produce not just development but also a demand on both state and municipalities to perform vital services for their electorate, or in other words, to make them accountable to the people. In that way, the improvement of the system of government and the development efforts will become more closely linked and pull in the same direction.