Civil war deeply disrupts and transforms policing. The two civil wars seriously undermined the ability of the states of Sierra Leone and Rwanda to maintain the rule of law (Baker, 2005b; 2006b). The state police were decimated during Sierra Leone's civil war 1991–20011 and virtually dissolved in Rwanda's civil war and subsequent genocide of 1994. Neither state has the resources to offer the protection from crime and the investigation of crime that their people seek. Sierra Leone can still only offer 8,000 police (6,000 operational) for its population of 5 million (that is 1:625) and Rwanda can only offer 5,500 for its population of 8 million (that is 1:1,454, c.f. England and Wales 1: 402). Neither of these state police forces are adequate in numbers, skills, resources or available national infrastructure to serve all citizens. They are confined to the centres of the main urban areas and to the main roads as they were before the wars. Ill-equipped and inaccessible to most, they are clearly not in a position to provide all the policing that is required by their citizens, nor can they replace long-standing local security arrangements and prevent new local initiatives. Even the most concerted externally led reform programmes are not going to change this reality in the short to medium term. This harsh reality begs the question of who then will provide policing to the poor of Africa? If the state cannot fulfil the core function of protection, then is there an alternative?
Post-conflict transition invariably sees contestation over the new order of policing. Immediately after the wars the new regimes and the local communities had to decide what sort of policing there should be ‘beyond the tarmac road’ and how far ‘beyond the state’ it should be allowed to be. The answers they reached in Sierra Leone and Rwanda, and the reasons behind those answers, are the subject of this article.
Policing, as understood in this article, is any organised activity that seeks to ensure the maintenance of communal order, security and peace through elements of prevention, deterrence, investigation of breaches, resolution and punishment. This differs little from the 1829 definition of Richard Mayne, upheld by the UK's Metropolitan Police on their web site: ‘the primary object of an efficient police is the prevention of crime: the next that of detection and punishment of offenders if crime is committed’.2 But though the Metropolitan Police assume policing is an activity only of the state police, African realities require the recognition that such activities can be authorised and provided by non-state groups. For the most part, I term the alternatives to state policing in Sierra Leone and Rwanda, local forms of policing, because they are authorised and provided by the local community (or part of it) for the local community (or part of it).
The article demonstrates that local forms of policing are widespread and preferred to state policing for reasons of effectiveness, accessibility and adherence to local definitions of what constitutes disorder. It also shows that their form and function varies widely. Though often called ‘non-state’ policing agencies, it will be shown that few are totally autonomous. Rather, the more typical form is a hybrid model where some form of relationship (formal or informal) with the state police exists. Put another way, it will become apparent that state power tends to infiltrate non-state policing.
Though local factors such as perceptions, values, poverty, cultural heterogeneity, geography and the presence of political entrepreneurs give a specificity to local policing (Baker, 2005a; 2005b; 2007a), from a national perspective, the country-specific pattern of local forms of policing will be shown to depend primarily on three factors: First, the nature of the conflict and peace settlement; second, the regime ideology (itself, of course, shaped by the experience of the conflict); third, the level of regime insecurity and fear of conflict recurring.
Policing is a broad cluster of functions, discussed in this article under the three headings of ‘crime prevention and intervention’, ‘investigation and resolution’, and ‘punishment’. First, however, the three key factors determining the reconstruction of policing in post-war Sierra Leone and Rwanda will be outlined.
Factors Shaping Local Forms of Policing
The Nature of the Conflict & Peace Settlement
Although both countries suffered a violent civil war, there are key divergences that have affected the willingness of local policing actors to work with the state in the post-conflict environment. In the case of Sierra Leone, the war saw atrocities committed by all sides (Gbere, 2005) and a confusing picture of changing allegiances on a temporary or permanent basis. With the withdrawal of police and many of the customary chiefs from the main conflict zones, army units (Sierra Leone Army, SLA), units of renegade factions (Armed Forces Revolutionary Council, AFRC), rebel units (Revolutionary United Front, RUF) and self-defence militias (Civilian Defence Forces, CDF) made the law. Most armed groups were predatory, though some, such as the CDF, acted for the most part as protectors of their communities and tried to promote order (Reno, 2007). After the war, the CDF, which despite its occasional atrocities still retained widespread support, was disbanded, whilst many customary courts remained closed. This left the state as the main security service deliverer. But failure to protect the population and the abuses of its armed forces in the war had done nothing to raise its credibility as a defender of the people. In this context there was little incentive for communities to work with the state in rebuilding policing.
The situation was quite different in Rwanda. It was not just a civil war that Rwanda suffered, but a genocide; one that saw one million citizens lose their lives in three months in 1994 (Prunier, 1995). For all the irregularities, there was a much clearer picture of who was on the side of justice and who was not. Hence those outraged by the horror of genocide needed little incentive to serve the regime that brought justice: they had either risked their lives in the Rwanda Patriotic Army (RPA) or suffered as victims until the RPA intervened. As a Police Commander said: ‘If you fought for injustice then your desire is to maintain it’ (Interview, 22 February 2006).
The two countries also differed in the outcome of the war. In the case of Sierra Leone, it was the existing regime that prevailed over the rebels; in Rwanda, it was the rebels who triumphed. The two different conclusions have shaped each regime's approach to policing. In Sierra Leone a political class returned to power that had been largely contained within the ‘fortress’ of Freetown and was used to relying on external advice and military support. This dependency on outside professional expertise had become deeply ingrained and was transferred to its policing policy. It uncritically followed a western policing model at the advice of western donors.
For its part, the Rwanda Patriotic Front (RPF) saw their success in the war and in the subsequent insurgency in the northwest, as due to winning the hearts and minds of the local people. As Cyprien Gatete, former RPA fighter and subsequent Assistant Commissioner of the Rwandan National Police put it:
We began as a very small group. We were poorly equipped and had no logistical support. We had to rely on the support of ‘family members’ … And they gave us food and cover and information about the movements of the enemy. We knew the population's support was crucial. In fact we were doing community policing! And it continued after we seized power (Interview, 16 February 2006).
Inevitably the new regime was keen to incorporate the participation of local people into its security and justice structures. Both regimes, therefore, looked to support in security to those who had assisted them in war.
A third difference was that the end of the war provided very different opportunities for the two regimes. Sierra Leone's police force was decimated, but not destroyed. Hence the regime had little choice but to seek to reform what remained. In other words, it passed up the opportunity to remove from the force human rights abusers and/or those accused of corruption. Rwanda, on the other hand, saw the ancien regime leaders and security agents largely swept into the Congo forests and beyond. It was closer to a clean slate than anywhere else and the opportunity was seized. So in all three respects the experience of war was distinct for the two regimes and this is reflected in the policing models they have pursued since.
The Regime Ideology
As the two regimes began reconstruction, the conclusions drawn (or not drawn) during the war had to be directed to specific policy choices: should it be a rebuilding of the old or the construction of something new; should it be the new or the past actors who authorise and provide policing; and should alternative policing systems be state planned and approved, or left alone?
The answer was of course partly dependent on the experience of war, as outlined above, but ideology played a part. The ideology of the Sierra Leonean leaders was state-centric and committed to formal rather than informal solutions. President Kabbah had been an international civil servant for almost two decades, holding a number of senior administrative positions at UNDP. The regime, therefore, accepted without question the donor position that policing was (beyond the withdrawal of the UN Mission in Sierra Leone, UNAMSIL) the prerogative of the state police and accepted donor funding that focused on state building (Gbla, 2006; Fitzgerald, 2004). It was an ideology that saw state building as the product of good governance, including western professional state policing. It never questioned that the state police should be the main provider (if not monopoly provider) of internal security. The regime wanted reform of the old forms, not new forms of policing. Hence though there were changes in senior personnel and management structures, the state police was to stay essentially the same. The war had taught the regime to defend the state institutions, not to overthrow them.
The Rwandan regime took a different approach to policing reconstruction. Even before the war, RPF leader Paul Kagame had embraced the centralisation and surveillance of the Frelimo government in Mozambique and enthused, along with his mentor Museveni, about popular justice in Tanzania and Mozambique (Museveni, 1986; 1997). In their eyes popular justice is popular in form because its language is open and accessible; popular in functioning because its proceedings involve active community participation; and popular in substance because judges are drawn from the people and give judgment in the interests of the people. Besides, it was evident that it would take time to build a police force from scratch and that even when the process was complete, resources would not allow for it to be adequate to provide all the policing needs of the nation. From the beginning Rwanda chose a policy of incorporating the participation of local people into its state security and justice structures. Such participation of the public alongside and in co-operation with the police was also seen as an instrument for healing both the aversion to the state police caused by the previous regime's oppressive and racist policing and positively for stimulating reconciliation and social cohesion through mutual cooperation.
Two different ideologies, therefore, have lead to two different models being pursued: the western policing model and a form of popular justice model (Baker, 2004).
The Level of Regime Insecurity & Fear of Conflict Recurring
Where the two regimes do agree is that the return to insecurity must be avoided at all costs. Civil wars make new regimes nervous of political opposition and renewed conflict. Hence neither of the two police forces have fully cast off their regime-policing role of harassment and suppression of political opposition.
The regime insecurity extended to their response to policing agencies other than the state police. Rwanda insists on strictly supervising all matters of security and very little space is allowed for security ‘beyond’ the state. Hence the voluntary community participation in policing is overseen by the local government structure and there are interventions to ensure ‘suitable’ candidates are chosen. Tens of thousands of unpaid local leaders keep watch over, manage and assist very small units of the population. Given that one of the local leadership roles is policing, it ensures that criminal and socially unacceptable behaviour is recorded and passed up the numerous layers of local government, the police and the central government, to the security office of the President. As a result, other informal policing has little chance of succeeding. Here then is a security structure that is nationwide and harnesses the voluntary assistance of thousands of local leaders to provide crime protection and investigation. Yet being in a centralised state this form of popular justice is under the guidance of the state. It is a case of the state infusing the informal.
Sierra Leone has no comparable informal structures to penetrate society and win compliance for official policies. The government has very little knowledge of what policing provision is going on beyond the main towns and only fleetingly attempts to employ the energies and knowledge of the local community or to supervise what the local community offers in the way of policing. Even its Police Partnership Boards, aimed at drawing in the assistance of citizens, give very little opportunity for genuine participation, often fail to meet and in one case those who attended a meeting to consider the Community Action Plan, found it had been drawn up by the police beforehand (Baker, 2005b, 2007). The insecurity of the Sierra Leone regime is expressed in their desire (along with the international community) to see no arms in the hands of any agency outside the army. All policing, including the state police (with the exception of the Operational Support Division) and commercial security companies, are prevented from possessing weapons. This, together with their emphasis on DDR (Demobilisation, Disarmament and Reintegration), is assumed (optimistically) to cut off the supply from would-be rebels.
Taken together, these three factors (the nature of the conflict and peace settlement, the regime ideology, and the level of regime insecurity and fear of conflict recurring) have shaped the profile of policing in the two countries and in large measure explain the different patterns between them.
Functions of Local Forms of Policing
This section discusses some of the local forms of policing groups active in the two countries and investigates the extent of their functions.
Crime Prevention/Intervention
Local forms of policing aimed at preventing crime are largely about patrols, particularly at night. Extended families and neighbours can and do keep their eyes out for strangers and unusual behaviour during the day, but at night with no street lighting, homes with outside kitchens, possibly no external fence/wall and with weak or no locks on the doors, they are more susceptible to crime. Further, it is more often at night that troubles arise with drunkenness, associated fights, prostitution and drug dealing.
What do local forms of policing offer in the way of protection? How are the patrols organised? Who authorises them and provides them? Are they effective? In the absence of the Sierra Leone Police (SLP) and sometimes with the failure of the customary structures (courts and chiefdom police) to re-establish themselves after the war, there emerged in Sierra Leone a clear security gap. Ironically, given their demonisation as violent combatants and idle unemployed civilians, it is the youth that frequently fill it. The more violent certainly perpetuate mob justice, however, in villages and poor townships many young people perceive themselves as ‘guardians of security’ (Interviews, Makeni, 20 February 2005).
In the town of Yengema in the east, where there is no police post, the youth stay alert at night. The Youth Chairman described the role of youth in crime prevention:
We know everybody and everything. If there is a fight they go to the youth group or after them the chiefs; only finally do they go to the police … We patrol at night. We respond if there is a problem. We harass anybody who brings drugs. We arrest them, destroy the drugs; we give them a beating. Solved! No more drugs! (Interview, 8 February 2005).
The town chief concurred that the area was ‘depending on youths to take care of us at night’, though he added that ‘youths are not always reliable’ (Interview, 8 February 2005). Similarly, in the nearby village of Tombodou, the chief spoke of youth controlling security: ‘they see [that] things go on normally’ and fill in for the inactive chiefdom police and absent SLP (Interview, 9 February 2005). According to the Tombodou women's leader, they ‘ensure local policing where SLP don't go. They make arrests and take them to the SLP’ (Interview, 9 February 2005). Youth also frequently provide unpaid security for alluvial diamond workings on behalf of the elders.
In parts of Freetown, youth groups participate in crime prevention and response. By the port, in high-density Krootown, Camp Divas Youth maintain a measure of order, especially among the youth. In the words of a member:
We never see the police down here … If there is fighting or stealing we take them to the police. And if there is provocation or abusive language we fine them! (Interview, 31 January 2005).
The Youth Secretary was asked how people responded when fights and thefts occurred: ‘People call the youth group; or the elders if there is a problem solving it. Finally they go to the police’ (Interview, 31 January 2005). A local tribal headman concurred that youth often intervened to stop fighting when the police failed to respond (Interview, 31 January 2005).
Another local initiative has been promoted by the Sierra Leone Police themselves. From 1999 police-community forums, known as Police Local Partnership Boards, have begun to be introduced. They are chaired by a civilian body that includes representatives of the significant groups and interests in the locality, including youth groups, chiefs, religious leaders and business persons (see Baker, 2007b). These Boards have improved communication between police and communities and provided intelligence, investigation, intervention, arrest and dispute resolution. On the negative side, they are elite dominated and most of the activities, initiatives and even finance are coming from the community. There is serious doubt about whether they are sustainable and will play any serious role in security provision in the short term.
Informal crime prevention groups exist in Rwanda, but they are not the product of spontaneity or local initiation, but responses to invitations of the state for communities to build on and extend the long-standing local forms of courts (gacaca). Here the local voluntary groups constitute, in effect, the lowest levels of local government, have responsibility amongst other things for the mobilisation and sensitisation of the local community in law and order matters. This not only means crime prevention in the form of patrols, but law enforcement; resolution and reconciliation mechanisms (individually or a gacaca); punishment for misbehaviour; making by-laws that reflect local needs; and the supervision of the local militia, the Local Defence Force (LDF). The structure has a significant ordering effect on social life and acts as the first line of protection against crime. Prior to reorganisation in February 2006, there were two levels below the lowest formal level of local government the elected résponsable, namely the nyumbakumi, who was in charge of 10–15 households and the chef de zone who was in charge of 200 households. Under the new arrangement the two have been merged to become umukuru and as a committee of four are responsible for 50 households or perhaps up to 200 in the towns. The résponsable is responsible, with a committee of seven and five LDF personnel, for about 500–1,000 households.
Local leaders work voluntarily and take their work seriously. Patrols are formed at the chef de zone level with each local nyumbakumi supplying two to five people and the LDF providing one armed member. They work under the responsibility of the résponsable. The frequency of patrols ranges from several times a week, to simply if and when there is a security problem. Should a patrol catch a petty criminal they are taken to the secteur (level above the résponsable) and locked up. For stealing bananas they may spend a single night, for more serious offences up to a week. After that there may be public shaming: ‘We then parade them in front of the local people before getting them to tell us all they know about other criminals in the area’ (Interview, 6 February 2006).
The regular night patrols were deemed by the population interviewed to be an effective deterrent and a useful means of stopping suspicious movement.3 But even watchful neighbours and night patrols cannot provide total protection from crime. Those who live by themselves can still be vulnerable, even during the day. A woman living alone in a high-density area in Kigali in a poor mud-brick house with no yard wall and only a flimsy door secured by a small padlock responded:
Thieves come a lot. Whenever I go for a short while I find things are missing. Every time I go to pray [at the church] or to the market I find things have gone: clothes taken [from the hedge] … Sometimes they push the door. Recently a thief took some clothes. We ran after the young man but he escaped. It is very common when I go to church. I keep going [during the service] to see if my door has been broken into (Interview, 30 January 2006).
Besides initiatives by youth and local leaders, many of the economic activities in Sierra Leone and Rwanda have a work-based association. The Motor Drivers Association in Sierra Leone, with a membership of about 5,000, control many of the commercial vehicle and mini-bus parking areas in the main towns. In Freetown they use car park attendants to oversee loading and protect passengers. In Waterloo, a small town east of Freetown, members take it in turn to be attendants. They have distinctive bibs and deal with problems relating to drivers, passengers and pickpockets. They exclude ‘undesirable’ persons. In Bo the drivers claimed responsibility for all vehicles in the town and took offenders to the police.
Something very similar takes place in Rwanda. ATRACO (Agence Rwandaise pour le Transportation en Commune, ATRACO-EXPRESS) is an association of about 3,000 taxi drivers that operates as a private company. It is responsible for the security of drivers, passengers and visitors in taxi parks throughout Rwanda. The taxis park in lines designated for particular destinations and each line has a ‘line leader’ in a green uniform responsible for that line. Every morning, the line leader is responsible to check the vehicles’ tyres and cleanliness. If members commit a disciplinary offence (for example, driving too fast or driving under the influence of alcohol or marijuana) they are first given a warning, then a fine, and finally the police are asked to take away their licence. Disputes and fights between drivers and between drivers and passengers are resolved by the line leader or if they fail to reach a solution the senior managers of ATRACO bring the parties together (Interview, 23 February 2006).
Investigation & Resolution
Investigation is the Achilles’ heel of all policing in Africa, state and community-based, owing to the heavy reliance on witnesses in the absence of investigative skills and forensic science facilities. Police CIDs have very limited resources and only a few commercial companies offer any serious services, mainly private investigative companies concerned with domestic and marriage disputes. Investigation by local forms of policing therefore, take the form of inquiry into the incident in the process of mediation and trial. Traditionally African justice systems outside the state structure have emphasised resolving disputes rather than seeking punishment. This applies to work-based associations such as markets and drivers, to customary structures, to restorative justice groups and to local government.
Almost everywhere in Sierra Leone, anti-social behaviour and its resolution are regulated by chiefs. Not that chiefs are always popular (Richards, 2005), but there may be very little in the way of attractive alternatives. One chief from the Bo area claimed:
People would not report directly to the SLP even if it was a criminal case such as wounding or theft cases. They come to the paramount chief first and they [the chiefs] give advice as to what to do. You see there is a general ignorance of legal matters (Interview, 3 March 2005).
A town chief in Tombodou, Kono District, made a similar claim:
As local chief I deal with 70–80 per cent of the work [disputes]. I see all cases of less than 30,000 Leones. People come to the Native Authority court because they have no respect for the authorities (Interview, 9 February 2005).
Using customary law, chiefs handle cases involving family law, debt repayment, inheritance, and land tenure (Fanthorpe, 2006). This system was the only form of legal system available before the war, and despite its decimation, it was the only one during the war. Since the war, for all the British support for maintaining the chieftain system and the money spent on reconstructing some of the burnt-out chiefs’ barrays (court houses), there have been difficulties in re-establishing the system and some courts only sit infrequently. For instance, of 20 in the Bo District, only 16 barrays were operating at the time of the research and there had been no salary for the Native Authority police since 1989. Further, many chiefs have also had their authority undermined because of their failure to protect the people and some lost their ‘mystique’ when they were seen being tortured and killed by the rebels or queuing with the people for food handouts.4 Fifteen young people in a northern village told me: ‘We don't send cases to the paramount … He hid in the bush with chiefs [in war]’ (Interview, 27 January 2005). Other complaints suggest that some chiefs are using their position to make money from the fines. Although chiefs insisted that they could only impose fines up to 10,000 leones, it is possible that some abuse their office.
In Rwanda, with the state's proscription of ethnic identity in favour of a single ‘Rwandan’ identity, it could not contemplate a customary system such as Sierra Leone's and any such identity is prohibited socially and in the courts – formal and informal. So although the responsibility for resolution still largely falls to the local leaders, these are either elected by the people (résponsable) or appointed by those elected (chef de zone and nyumbakumi). The latter two constitute the lowest and informal levels of local government. Either personally with the disputants or in gacaca, they seek to resolve the problem rather than have it enter the formal state legal system. They can legally deal with anything that concerns property worth up to three million francs (about £3,000). One chef de zone in Kigali summarised the desire for resolution that motivates the system:
I try to make relationships between me and the population and for them to see that we are one and are able to solve our problems together. A chef de zone makes sure that the population don't see him as a leader among them, but one who is there to solve their problems. Before they go to the police they come to us. Whenever they bring their problems we speak to them as family members not as the police (Interview, 27 January 2006).
If a crime is committed, complainants have to follow the local leadership hierarchy. This means beginning with the nyumbakumi. A nyumbakumi in Kigali described his role:
I am responsible for 10 houses: seeing how people live, things they do; to know what problems they meet with, see if there are any disputes or conflicts between people … I try to reconcile any problem (Interview, 28 January 2006).
If that fails, people take it to the next level, the chef de zone.
I call seven or so to a gacaca court. I listen to what they say [their assessment] and decide [the verdict] on the basis of what they say. If the accused man accepts he did it, then it is agreed the payment he has to pay and the instalments (Interview, 20 January 2006).
If the chef de zone fails to resolve the dispute it goes to the next level, the résponsable. According to two résponsable:
The résponsable calls his advisers and together they make a decision. But it is not a court that can punish; it tries to sort out problems. At the end the résponsable may tell the ['guilty'] party to buy a crate of Fantas for 1000 francs and then there are hugs and forgiveness. If it doesn't work out then it goes to the secteur level which is a court with power. It tries to bring reconciliation but they may use the law. If it can't get people to sort it out it goes to the official court of the commune [next level] (Interview, 6 February 2006).
Wherever there is a problem I call a gacaca. All 9 of the court will sort out land disputes for instance. We prefer to sort out disputes amongst ourselves rather than forward them to outside (Interview, 13 February 2006).
The emphasis therefore is on resolving disputes in the time-honoured and still almost universally supported manner of gacaca, and to prevent the individuals and community from having to go to the slow, and potentially community-divisive process of formal courts where there are winners and losers. This concern for community stability and order by both Rwandan gacaca and Sierra Leonean customary courts is a much stronger motivation than any pecuniary gains from fines.
Punishment
Local forms of policing are frequently accused of human rights abuses, especially when it comes to punishment of those found guilty of crime. However, corporal punishment, though a regular feature of mob justice, is not the only or even the main punishment used in local forms of policing in Sierra Leone and Rwanda. Conflict resolution is preferred, although dispute resolution processes (whether work-based association meetings or customary courts) may still issue punishments, fines or exclusions from the work place. There are times when beatings are administered and this is invariably with the support of most of the local community. For example, the Camp Divas Youth of Freetown maintain that they ‘flog’ all those who fail to pay fines imposed for bad behaviour. Market traders in Freetown admitted that if a thief was caught he was usually beaten on the spot. Likewise in Makeni, Sierra Leone, the drivers association alleged that for dangerous driving by members: ‘we give them lashes’ (Interviews, 20 February 2005).
Corporal punishment does not play a large part in the local government system of Rwanda, though it is not totally absent. If a résponsable finds someone involved in crime they are usually taken to the secteur and locked up for a few days. When they are released, as mentioned before, they may be paraded in front of the local people, have to ask them for forgiveness and do community service in the form of joining the local patrol. They may also be required to provide information on other criminals and their organisation. For the most part corporal punishment is avoided although one résponsable admitted that though criminals cannot be beaten, ‘if he runs away then you are allowed to throw stones. The Local Defence are allowed to shoot at serious criminals’ (Interview, 6 February 2006).
Thus, both Sierra Leone's spontaneous and Rwanda's supervised informal policing function very similarly in terms of responses to crime, both following long-held patterns of dispute resolution.
Attitude to State Police
While it is often assumed that Africans want more state police (Goldsmith, 2003), the cases of Sierra Leone and Rwanda with the ubiquity and popularity of the local forms of policing show that it is not the case that people are missing the presence of the state police. Asked if they would like to see more of the police in their northern Sierra Leonean village, the villagers said they already had a local court ‘for conflicts’ between men and women: ‘There is no need for a post … What do we need a post for? … We are all right’ (Interview, 20 February 2005).
In Rwanda the police are rarely seen beyond the tarmac road unless they are called to a serious incident. But far from people feeling abandoned, most of those interviewed felt that their local security structures offered them what they needed. Typical was the view of a farmer living in a poor district in the east who claimed that the community rarely had more than the odd petty thief and that they were well organised in terms of patrols and restorative justice by the nyumbakumi. When asked if the community ever saw the police, he replied: ‘No they never appear in these areas unless there is a serious crime that has happened’ (Interview, 14 February 2006).
It is evident, therefore, that in the absence of the state police, local forms of policing in Rwanda and Sierra Leone undertake a wide range of services, including prevention, intervention, investigation, resolution and punishment. Few in the two countries are far from access to these local policing services.
How Autonomous are Local Forms of Policing?
In countries where the state police have never provided a nationwide service for all, even before their partial urban service was further weakened by war, there have always been alternative providers. Today in Sierra Leone and Rwanda they are a diverse group. It includes in Sierra Leone youth groups, customary chiefs, community peace monitors, market vendors’ committees, taxi drivers’ associations, commercial security guards, mine monitors, beach police and community police forums. In Rwanda there are informal local government structures, the LDF, commercial security and work-based security associations. One of the most interesting questions concerns how these groups relate to the state: what degrees of autonomy do they have if any; is the relationship competitive or co-operative; beneficial to both parties or only one party; who sets the terms of the relationship? In terms of the boundary between state and non-state, is this distinct, blurred, fixed or shifting?
Elsewhere I have noted that in Africa there are degrees of distance or escape from the state's gravitational pull (Baker, 2006). Some groups pay no deference to the state, but others are emerging within the boundaries of state initiation or at least state influence and approval. There are different levels at which the state may shape and influence policing agencies other than the state police. Sponsorship, regulation/ criminalisation, networking/exclusion, incorporation and training, are all techniques by which the state may seek to maintain a level of control over ostensibly ‘non-state’ policing, and seek to uphold its sovereignty. From the state's point of view, a relationship provides an opportunity to infiltrate non-state policing. From the point of view of the local forms of policing, the relationship provides a measure of official legitimacy.
In the case of Sierra Leone and Rwanda few policing agencies are totally autonomous. The closest case would be the youth groups in Sierra Leone who have little to do with the police for the simple fact that they provide local security in areas where the SLP are rarely seen. They are prepared to break the law, especially in terms of violence meted out, to achieve their goals of protection and investigation (or sometimes trials and sentencing). Such informal organised security groups whose emphasis is on punishment often bear the pejorative name of vigilantes. Yet in the case of Sierra Leone they are not averse to working with the state police when it suits. Outside a police divisional headquarters in Freetown I watched 40 youths demonstrating loudly for compensation for a cow they had ‘arrested’ wandering down the main road. Inside the station the Local Unit Commander negotiated with the youth and the cow's owner, until a financial settlement was reached and the crowd left in jubilation. The ambivalence of the relationship is shared by the police. There is no doubt they know of their existence and acknowledge that they do at times bring to the police stations those accused of crime. Yet the police are wary of youths in general. For example, the Annual Crime Report of the SLP attributes the increase in crime rates in the Freetown area to, inter alia, ‘lots of idlers and unemployed youth roaming about’ (SLP, 2005), while one local police commander described the youth representatives on the Neighbourhood Watch as ‘criminals’ (Interview, 20 February 2005). Despite these strained relations with the police, the youth enjoy a considerable amount of popular local legitimacy on account of the approval of their elders, as has been shown above and few in their communities would harbour doubts about the use of violence. It demonstrates that the dividing line that is seen in the west between positive and negative local forms of policing is less clear on the ground.
The more typical form of policing, however, is a ‘hybrid’ model where some form of relationship with the state police exists (Johnston, 1992). The relationship can be relatively tight. Rwanda's local community structures, for instance, are almost a case of incorporation. Because these structures are accessible, free, offer a quick response by local people that are trusted, and are effective in protection, investigation and reconciliation, they are well regarded by the people. Officially they are outside the legally constituted local government system, but they link smoothly with the latter and pass on to them information and unresolved cases. For the most part they are free to handle their cases in their own way and only turn to the police if the crime is serious. Not only do the police approve of this free nationwide service of first response to crime and disorder, but they draw from it substantial intelligence. The local leaders each keep a book of local crime and disorder and of people known for different trouble. That information goes up the hierarchy until it reaches the commune, where it is passed onto the local police commander. Said a chef de zone: ‘I give reports [to the résponsable] of people who drink, fight, exceed sleeping time!’ (Interview, 27 January 2006). And a résponsable said: ‘Every Wednesday we give a written report of our area’ about those ‘who have been fighting, stealing’ (Interview, 16 February 2006). Though it is the state that sets the terms of the relationship, both sides regard the relationship as beneficial.
Some of the work-based security associations have looser relationships of cooperation with the police. The motorbike taxi association of Rwanda, when asked how it protected its members from robbery answered: ‘The police help a lot. We have the authority to call them. They come quickly’ (Interview, 30 January 2006). At the same time it appeared that the police used them as an organisation to provide information to track down criminals.
A still looser alliance is found between the police and the customary courts of Sierra Leone. The system is structured such that there are chiefdoms with their sections and sub-section, and every town outside of Freetown has a chief. Cases are reported from the town chief's court to the section chief's court, then to the paramount chief's court. Their courts would probably be left entirely to their own discretion, were it not for two factors. First, there are rising complaints by users and human rights NGOs against abuses in customary courts. Allegations include discrimination, particularly against women (chiefs, for example giving minimal fines for or ignoring cases of wife beating); illegal detention; charging excessive fines for minor offences; and the adjudication of criminal cases contrary to the constitution. These require some token of state and police response. At the same time, the second factor militates against too much interference, namely that chiefs are an essential part of neopatrimonial rule, particularly in their role of mobilising support for the ruling party. It is not an accident that the government has refused to bring any reform to the customary policing and courts since the war, as it needs to maintain chief loyalty. Yet there is a link between the two legal systems. If the paramount chief is unable to resolve a case, it goes to the magistrates. In addition, the chairman of the customary court is appointed by the Minster of Local Government, and despite the complaints, the police in practice rarely intervene. In their view: ‘Our priority is the main towns. We cannot cover all the villages. I am confident that the native authority police can cope, except with the mining youth’ (Interview, 21 February 2005).
The relationships in the two countries between local forms of policing and the police vary widely, but in no case could it be said to be competitive in the way it is sometimes between commercial groups and the police (Baker, 2005a). Yet despite the existence of relationships between state and local forms of policing, there is little actual mutual exchange of intelligence or joint operations, even in the case of community policing forums which are ostensibly set up to facilitate this. The intelligence traffic tends to be one-way and police operations are normally exclusive, reflecting the fact that in both countries it is the police who are setting the terms of the relationship. Even in a weak state like Sierra Leone, local forms of policing are not often beyond the reach of the state. Having said that, this does not mean it is within the state's arena of trust. The role of local forms of policing, from the point of view of the police, is that they are essentially there to complement them rather than to be partners with them.
It is apparent, therefore, that as discussed in the introduction to this Special Issue, the boundary between the public and the private, or in this case the state and local forms of policing, is complex. As elsewhere in Africa, the boundary is porous in terms of the crossing of information and cases handled. Transgression may be formal or informal (based on personal relationships), but it is clear that such relationships make the usage of the terms state and non-state problematic. The boundary is constantly negotiated – with the police agreeing, for example, to keep away from markets and taxi ranks or turning a blind eye to mob and vigilante violence. The boundary could also be said to be shifting as local groups formerly regarded as non-state are incorporated into officially approved community police forums and there is always the possibility of the role of the chiefs in policing moving from the legal to the illegal with new legislation. No boundary can be regarded as beyond renegotiation (Wood & Dupont, 2006).
Importantly the boundary does not correspond with the boundaries proclaimed by the state. States and their police are constantly constructing an image of policing and its boundaries that does not correspond with reality. State monopolies of force and public/private distinctions are largely fictions constructed by the state. Such statements present an image of the state as the sole legitimate and effective provider of nationwide policing, while non-state policing is typically portrayed as being beyond the boundary of the formal and legal, and as largely unnecessary, biased, abusive and illegal. Links across the public-private boundary are seen as problematic except in very limited circumstances. Although state policing retains political and symbolic significance, in practice local forms of policing are the policing of first choice for most Africans in most circumstances and there are (for reasons of mutual benefit) widespread links between public and private policing agents, especially at the local level. This divergence between practice and image is subject to an ongoing power struggle and negotiation such that the boundaries between public and private, legal and illegal, state and nonstate are subject to change and reconstruction (Migdal and Schlichte, 2005).
In addition to blurred boundaries some have spoken in terms of loose networks that bring together the nodes of policing providers, though with the state police as the primary node (Abrahamsen & Williams, 2007; Wood & Dupont, 2006; Loader, 2000; Newburn, 2001). There are certainly elements of that in the cases of Rwanda and Sierra Leone, but bearing in mind that the links are often, as in the rest of Africa, limited in time, content and may rely on personal informal relationships, the relationships are better conceptualised by Johnston and Shearing's (2003) ‘security governance’ in which no prior assumption is made of state dominance, or by Wood and Shearing's ‘nodal governance’ (2007) which does not assume that nodes are necessarily networked.
Can We Live With Them?
Beyond the tarmac road where the majority of the people of Rwanda and Sierra Leone live, policing will continue as it has for centuries to be in the hands of local actors. Neither violent war, nor a centralising state, nor the international donor community will alter this reality, though they may shape the particular manifestations of it. Yet none would say that working with them will be easy. They cannot be romanticised into the answer for the failure of the state police. The world of local forms of policing is complex, diverse and dynamic, making a common strategy problematic and not one readily programmable. In addition, the resistance of local and national elites to the building up of institutions outside the state is inevitable and the police, too, are inclined to see alternative policing groups as rivals that threaten their professional monopoly. It is certainly true that anything run by volunteers is always prone to being unreliable and unsustainable. Further, such local groups are particularly susceptible to corruption and to the abuse of power and manipulation by local elites, although this is of course also a problem with large national institutions like the state police. Undoubtedly local police activities are not fully compliant with international human rights standards and despite increasing awareness, discrimination, especially against women and the poor, is still common. However, when these defects are acknowledged in the state police, the usual reaction is to devise methods of reform and there is no evidence that local forms of policing are inherently more resistant to reform than the state police. Sceptics also point to irregular record keeping, though this is often a reflection of the same poverty that makes it problematic for the police service too. Again, the evidence does not suggest that local policing forms are unwilling to document cases.
Undoubtedly there are considerable difficulties in working with local forms of policing. Yet it has to be acknowledged that they offer intriguing potential as development partners. After all, there is not a village, town or city slum where such institutions are not already in place offering some form of service delivery. These groups can be highly effective, since they have local knowledge and significant local support. Being locally owned they also overcome the problem of resistance to external reforms. And though war and conflict quickly decimate state structures, these groups have a history of being resilient and of being adaptable to the changing needs of the local populace. The fact that they are not in need of large capital expenditure means that their services could be improved in manner that is both affordable and sustainable. While for some critics, the difficulties of building links between the state and local forms of policing may seem insurmountable, the fact is that those linkages often exist in informal if not formal ways. As such, any programme that seeks to improve security for the poor, may thus want start by strengthening these existing linkages of information exchange and delegated activities.
The price of order in poor post-conflict states that must rely on local forms of policing for most of their inhabitants most of the time may well be a deficit in terms of justice. Local forms of policing may be flawed in the eyes of both users, government and donors; but in a context of less than fair and accountable state policing, they are not to be dismissed lightly. As this article has argued, they are not without support and a modicum of accountability and, most positive of all, most are capable of being upgraded – and would welcome that. Of course there are those groups that act like criminal protection rackets and are answerable to no one. No one would support their existence, including most of the communities that they prey on. Fortunately they constitute only a very small proportion of groups in Sierra Leone and Rwanda. For the rest, they can be reformed to enhance their service delivery quite as much as the state police. If policing is to be provided for the poor of Africa, then it will not be without the help of local forms of policing. States need their partnership. They deserve state support.