40470 Aceh Conflict Monitoring Update 1st ­ 30th November 2006 World Bank/DSF As part of an analytical support program to the peace process, the Conflict and Development Program, within the World Bank Office Indonesia, is using a newspaper conflict mapping methodology to record and categorize all incidents of conflict in Aceh as reported in two provincial newspapers (Serambi and Aceh Kita). The Program publishes monthly updates and analysis of the data, complemented by fieldwork where possible, in both English and Indonesian.1 While no GAM-GoI conflict cases occurred this month, a variety of cleavages from the thirty- year conflict continue to manifest in violent incidents and tensions. In this Update we present a tentative analysis that maps some of the key pre-MoU cleavages onto more recent conflict cases, giving examples from incidents in recent months. It shows that the forms of local conflict arising are not divorced from Aceh's past. Rather, disputes over aid, political disputes and vigilante acts are often a consequence of the ways in which the conflict affected social and state-society relations. This month saw the highest number of violent local level conflict incidents on record since we began monitoring at the beginning of 2005. 18 cases were recorded, double the previous high, which was recorded last month. This increase was primarily related to pilkada (local elections)- related incidents (8 cases) and acts of vigilantism (7 cases). Local level conflict increased by 71% this month to 82 cases. This brings local level conflict to levels comparable with the period from February to August this year, suggesting that the drop over the past two months was an anomaly related to Ramadan and the papers' focus on (then non-conflictual) early pilkada news. The recent rise in violent local conflict has important implications for those seeking to consolidate peace in Aceh. The analysis suggests that many of the current manifestations of earlier cleavages relate to reintegration processes, and particularly the delivery, or lack thereof, of reintegration assistance. Improving the delivery of reintegration assistance is probably the priority for ensuring past conflict cleavages do not lead to violence and are redirected into more manageable forms of conflict. There is a need to continue to focus on finding ways to improve trust between groups that were involved in the conflict: between GAM and anti-separatist groups; between security forces and the community; and between individuals and groups who had different loyalties during the conflict. Dealing with issues relating to the fragmentation of authority within GAM structures is key. Further, increased attention must be given to improving the ability of local government, communities and the police to prevent and respond to criminal acts. No GAM-GoI incidents reported in November, but post-conflict tensions remain Once again, no direct GAM-GoI conflict incidents were reported in November (see Figure 1). Indicative of the continuing strength of the Helsinki peace process, the participation of GAM- affiliated candidates as independent candidates in the pilkada elections has not, to date, inflamed GAM-GoI tensions. 1 There are limitations to using newspapers to map conflict; for more analysis of the methodology see: Patrick Barron and Joanne Sharpe (2005). "Counting Conflict: Using Newspaper Reports to Understand Violence in Indonesia", Indonesian Social Development Paper No. 7. Jakarta: World Bank. This report, as well as the monthly monitoring updates, is available online at: www.conflictanddevelopment.org. The dataset is available for those interested; please contact Blair Palmer at bpalmer@worldbank.org. 1 Figure 1: GAM-GoI and local level conflicts by month GAM-GoI Local Level Conflict 90 80 70 60 50 40 30 20 10 0 JanFebMarAprMayJun Jul AugSep OctNovDecJanFeb MarAprMayJuneJulyAugSep OctNov 05 MoU 06 Nevertheless, tensions and cleavages built up through years of conflict do not disappear instantly. Indeed, the form and frequency of current conflict incidents continue to be influenced by the legacy of the violent vertical conflict in Aceh. While the MoU and peace process have resulted in an end to direct confrontations between GAM and the security forces, post-conflict tensions stemming from previous conflict cleavages are apparent and increasingly resulting in violence. Table 1 presents a preliminary analysis of some ways in which five key pre-MoU cleavages are manifesting in contemporary conflict incidents.2 Tensions are apparent at numerous levels, from high politics in Banda Aceh to localized community tensions. In most cases, these are not of a form that could seriously undermine the peace process. However, it is important to address some of the grievances that underlie such cleavages. This will necessitate formalizing processes to deal with (real and perceived) atrocities from the conflict period. It will also mean taking steps to ensure that small localized incidents are not grouped together and politicized as a basis for mobilization by elites with broader political goals. I. GAM vs. GoI tensions While there have been few incidents of direct confrontation between GAM and GoI this year, GAM/GoI tensions show through in a number of contemporary conflict forms. Residual tensions both between individuals and between institutions can lead otherwise small grievances to be acted on, leading to violent vigilante attacks. These tensions may have been a factor in GAM/KPA's exit from BRA, after the two sides were unable to form a positive working relationship on reintegration issues. Security forces and former combatants are also in competition over control of various illegal economic activities, such as the marijuana trade. Reports from the field indicate that in Pidie, many former combatants are now involved in selling marijuana. Past GAM/GoI tensions could also harden affiliations with either national parties or local candidates leading to increased political conflict. As an example of an incident in which the GAM/GoI cleavage seems important, on November 9th, Serambi reported that the KPA office in Lhoksukon, Aceh Utara, was vandalized by about a dozen armed police officers, and an office employee was allegedly assaulted. Two hours earlier, villagers, including KPA members, had raided the rented room of a police officer, suspecting that khalwat (suspected sexual contact between non-married individuals) offences often took place there. 2Note that this is a tentative analysis, and feedback is welcome. 2 II. GAM vs. anti-separatist groups These tensions have resurfaced in disputes over reintegration benefits, notably in Bener Meriah and Aceh Barat Daya. Some GAM/KPA members have voiced frustration that reintegration benefits to anti-separatist groups have flowed quicker than those to former GAM combatants. In addition, this cleavage mapped onto splits between candidates in the district election in Bener Meriah, leading to tensions in the lead-up to the election. Table 1: Pre-MoU cleavages manifesting in post-MoU conflict incidents Cleavage Contemporary Conflicts Examples I. GAM vs. 1. Individual (not institutional) revenge · Payu Bakong incident - GAM/TNI GoI and animosity between security shootout following a beating (July 2006) personnel and combatants · Various incidents handled by Aceh Utara 2. Conflicts over reintegration benefits DiCoSa (April 2006) (conflicts between BRA and · GAM withdrawal from BRA (June 2006) KPA/GAM about program design) · Cases related to control of rent-taking on 3. Challenges to local government east and west coast roads (World Bank authority study on illegal payments in trucking - 4. Conflicts over control of illegal March 2006) economy (marijuana, illegal · Former combatants moving into illegal logging, extortion, protection) logging (cases found in fieldwork in 5. Political conflicts: national parties December 2006) vs. local candidates II. GAM vs. 1. Conflicts over reintegration benefits · Protests in Bener Meriah by anti- anti-separatist (conflicts about who is eligible for separatists about reintegration funds groups what benefits) (April 2006) 2. Political competition in local · High tensions in Bener Meriah elections (November 2006) when two anti- separatists and one GAM commander launched candidacies (cited in Pilkada Update) · Tensions in Aceh Barat Daya, over reintegration funds, and attack by anti- separatists on SIRA office (Feb 2006) III. Intra- 1. Political competition in local · Attack on Humam Hamid (November GAM3 elections 2006) 2. Disputes over allocation of · Other incidents related to Human-Irwandi reintegration funds split · Split in GAM in Aceh Barat Daya over use of reintegration funds IV. 1. Vigilante attacks on security forces · Two vigilante attacks in October 2006 Community 2. Tensions related to post-conflict · Peudawa incident ­ police kill civilian, vs. security trauma leads to demonstrations (March 2006) forces · Local tensions compounded by high levels of trauma (IOM report on psychosocial trauma - September 2006) V. Intra- 1. Vigilante attacks · Killing of keuchik (village head) in Aceh community 2. Political conflicts in Selatan in September (GAM vs. GoI villages/leadership challenges · IDP return problems in Bener Meriah loyalties) 3. Problems with IDP returns (December 2005) 3There are well known tensions between older Sweden-based GAM and younger Aceh-based GAM, but in the past there has also been a leadership split between GAM and Majelis Pemerintahan GAM (MP-GAM), as well as various horizontal tensions within GAM. See Kirsten E. Schulze (2004), The Free Aceh Movement (GAM): Anatomy of a Separatist Organization. Washington: East-West Center, pp. 19-24. 3 III. Intra-GAM divisions The political split in GAM, which led to two candidate pairs with affiliations to GAM standing in the gubernatorial elections, is well known. Other internal divisions and enmities within GAM are less discussed but have the potential to be expressed as GAM transforms from a military organization to a local political party. Disputes over how reintegration funds are distributed through GAM structure are relevant here. KPA leadership has had to make decisions on who should be included on the MoU list of 3,000 GAM combatants as beneficiaries for reintegration funds. Frustrations over the slow pace of funds reaching the ground has also led to discontent from below, and raises the possibility of local KPA leaders breaking with those higher up the chain of command. IV. Security forces vs. community There are high levels of trauma in Acehnese communities due to abuse suffered at the hands of the security forces. This needs serious attention by donors and government. On several recent occasions, community members have unleashed violence on members of the security forces in vigilante attacks. V. Intra-community tensions Tensions between those who supported GAM or GoI during the conflict are present in many communities. The tensions have largely not led to explicit reprisals, but it is easy for such tensions to erupt when some unrelated grievance emerges. For instance, there have been several recent vigilante attacks against village leaders. Further, the return of conflict IDPs is a sensitive issue, and often provides a basis for previous tensions to reemerge. Tensions between former combatants and victims of the conflict, and between other groups in the community, can also manifest in disputes involving programs such as BRA-KDP (see Box 1). Box 1: BRA-KDP Assistance for Conflict Victims BRA is channeling over US$60 million in resources through the Kecamatan Development Project's (KDP) community-based mechanism to assist victims of the conflict over the next two years. The 2006 round of the program will cover over 1,700 villages in 67 kecamatan. The program involves block grants being given to participating villages, based on population size and history of conflict. Village meetings are then held to decide how funds should be divided: to individuals, to groups, or for community-wide projects. Introducing local collective decision-making on usage of funds raises the potential for conflict. Previous research has shown that KDP mechanisms are generally strong enough to handle tensions between different groups over how funds should be allocated; KDP has operated since 1998, including in many `black areas' where conflict raged. Initial supervision trips for BRA-KDP have found that in some places post-conflict tensions are upsetting the social cohesion necessary for the program to work. Former combatants are angry that they cannot be directly included as beneficiaries (they do have access to the funds as community members, but have no priority by virtue of being former combatants; there is separate funding for them as combatants but its distribution has been delayed). The `mapping' process for deciding who is a victim and hence eligible for funds is often contentious. So far, problems have generally been small and have been mitigated by skillful facilitation from KDP staff and local leaders. Yet the program does demonstrate how post-conflict cleavages can potentially impact on the performance of development projects. In order to reduce these tensions, mechanisms are now in place to strengthen the relationship between BRA and KPA, to improve socialization of the BRA-KDP program as well as reintegration assistance for former combatants, and to improve complaints handling procedures. Additional training for program staff and intensive monitoring of the program will also take place, and a program of evaluation will be conducted. 4 Pre-MoU cleavages are perhaps most evident in tensions surrounding livelihood issues for former combatants, including BRA assistance programs. Unfulfilled livelihood needs of former combatants have been flagged in previous Updates as helping lead to extortion, vigilante attacks, and a rise in armed robberies. Improving the delivery of reintegration assistance is probably the priority for ensuring past conflict cleavages are redirected into more manageable forms of conflict. Better delivery of reintegration assistance will become a key test of the new provincial government, one which has serious implications for the security situation in Aceh. At the same time, there is a need to develop activities to improve trust and social relations between groups who were previously on opposite sides of the conflict--between GAM and anti-separatist groups; between security forces and the community; and between individuals and groups who had different loyalties during the conflict. In the short-run, it will also be necessary to ensure that GAM/KPA structures remain cohesive to avoid splits within the organization and a resultant loss of social control over former combatants at the local level. Attention must also be paid to improving conflict resolution--by police, local government, and by communities themselves--to prevent small incidents of crime escalating into larger problems. Pilkada-related incidents increase sharply but no large scale incidents This month the number of election-related incidents jumped drastically from zero to 12 violent cases (see Figure 2), coinciding with the start of the official campaigning period on November 24th. However, former conflict-related cleavages (such as divisions along ethnic or anti- separatist/GAM lines) have not re-emerged in election-related disputes, showing the extent to which commitment to the peace process is strong from both parties to the MoU, as well as other groups in the community. Instead, forms of local tensions and election dynamics seem to vary widely between districts. The majority of elections-related incidents relate to turf scuffles between campaign teams rather than mass incidents between supporters.4 Figure 2: Pilkada-related conflict incidents by fortnight Non-violent Violent 30 25 20 15 10 5 0 Jul_1 Jul_2 Aug_1 Aug_2 Sep_1 Sep_2 Oct_1 Oct_2 Nov_1 Nov_2 Jul_1 is first half of July, Jul_2 is second half, and so on Source: UNDP/WB pilkada newspaper dataset Local level conflicts increase dramatically After last month's low level of local conflict (likely due to the Ramadan fasting month as well as the newspapers focusing on pilkada-related stories to the exclusion of local level conflict incidents), this month the number of local level conflicts jumped to 82, an increase of 71% (see Figure 1). This jump is largely related to the lead-up to the local elections. Aid, however, remained an important issue in 15 conflict incidents this month, with many such conflicts focused on problems with contractors, corruption allegations, or the slow pace of aid. 4 For more information on pilkada-related dynamics, see the two Aceh Pilkada Dynamics Updates, available at: www.conflictanddevelopment.org. 5 Aid-related conflict remains important Last month 42% of conflicts were aid-related; this month that figure fell to 18%. However, this is largely due to the high number of pilkada-related incidents. The actual number of aid-related conflicts dropped only marginally, from 20 to 15 (see Figure 3), and the breakdown of the issues highlighted in these conflicts remained similar to last month (see Figure 4). Notably, several conflicts were related to a "lack of transparency". Citizens are bringing forward complaints without concrete evidence of corruption, but on the basis that the funds were not being handled transparently. This is a promising development. An empowered citizenry that can demand transparency from office holders and development practitioners is necessary for anti-corruption strategies to be effective. Figure 3: Aid-related conflicts by month Figure 4: Breakdown of aid-related conflicts for November Aid-related Conflicts Local Level Conflict 90 contractor problems 80 20% corruption 70 allegations 60 34% 50 land 40 ownership issues 30 13% 20 10 0 lack of JanFebMarAprMayJun Jul AugSepOctNovDecJanFebMarAprMayJuneJulyAugSepOctNov transparency late aid 05 MoU 06 13% 20% Land issues remain important. On November 1st, Serambi reported that the Simpang Keuramat Community Health Center, in Aceh Utara, was shut down by villagers claiming it was built on their land without due compensation being paid. During November, there were a number of incidents involving the closure of the Banda Aceh to Calang road. Protesters claimed that the road was built on their land without compensation. At points the road was closed; at other times they allowed vehicles to pass after paying a toll. For more on connections between aid and conflict, see Box 2. Violent incidents double This month saw 18 violent incidents recorded, double what was recorded for last month and in fact the highest monthly total of violent incidents in our database (see Figure 5). These incidents can be separated into three major types: those related to pilkada (8), vigilante incidents (7), and other incidents (3). Pilkada cases can be further divided into those involving only property damage (4) and those involving violence against people (4 cases). Vigilante cases are described in more detail below. The forms of local violent conflict in November are presented in Figure 6. Figure 5: Violent and non-violent local level conflicts by month Figure 6: Breakdown of violent incidents for November Violent Local Level Conflict Non-violent Local Level Conflict "Moral 90 Pilkada-related vigilante" 80 violent attacks attacks 28% 70 22% 60 50 40 "Retribution" 30 vigilante 20 Pilkada-related incidents 10 property 11% damage 0 22% Other violence JanFebMar AprMayJun Jul AugSepOctNovDecJanFebMarAprMayJuneJulyAugSepOctNov 17% 05 MoU 06 6 Box 2: Tsunami Aid and Conflict The World Bank is currently conducting a study of connections between conflict and tsunami assistance. The implementation of tsunami programs is often affected by the conflict dynamics that predate the tsunami within that community. For example, in many conflict-affected areas villagers continue to be reticent to engage in decision- making relating to village development, due to trauma from the conflict period. This affects tsunami programming, since it leaves decisions to local officials, rendering programs more susceptible to capture by local elites, and making it less likely that vulnerable groups (including conflict IDPs and widows) will actively participate. The interaction between tsunami aid and conflict goes two ways; tsunami assistance, in turn, can aggravate pre- existing tensions--both those between villagers and those between villagers and village officials--thus leading to an increase in conflict (as has been shown in these Updates). Another source of tension is the differing amounts of aid received by tsunami victims and conflict victims. In housing reconstruction, BRR allocates US$ 5-6,000 per house, and houses are also built through contractors or community-based mechanisms that provide a system to support beneficiaries. In contrast, BRA allocates about US$ 3,500 per house for beneficiaries whose houses were burnt during the conflict. These funds are generally transferred to individuals without any institutional support, which leaves the funds more susceptible to bureaucratic manipulation, and leaves the beneficiaries left to build their houses themselves. In villages which were GAM strongholds, former combatants, and in particular local KPA leaders, often play an important unofficial role in monitoring or overseeing development programs. The extent and impacts of the role played by KPA leaders in development projects varies, but the unofficial nature of their involvement means that it is difficult to hold them accountable. In some areas KPA people also appear to play an important role in the supply of construction material, including providing transportation and security to sub-contractors. Organizations offering tsunami assistance need to be aware of conflict dynamics in their target areas, and of how their aid programs interact with those dynamics. Where possible, programs should be focused on overall needs (whether they are a result of the tsunami, the conflict or other factors) rather than on the cause of those needs. Vigilante incidents increase sharply, with many "moral vigilante" attacks There was a sharp increase in vigilante incidents, from two last month to eight this month. All but one of the November cases involved violence. These incidents are of several types. "Moral vigilante" attacks are usually beatings for khalwat (suspected sexual contact between non-married individuals) offences. The WH (Wilayatul Hisbah, or "syari'ah police") may or may not be involved, but victims of these attacks are usually turned over to the WH or the police following their beating. Four violent "moral vigilante" attacks were reported this month. In addition, a male foreigner was arrested for a khalwat offence in Meulaboh, but was not beaten.5 A more rare type of "moral vigilante" attack also occurred this month when a religious institution was attacked for inappropriate teachings (see Box 3). Another type of vigilante incident is "retribution" attacks, in response to some specific infraction committed by another group. The attack on the KPA office described earlier is one example of this; see Box 3 for another. Retribution attacks are usually underpinned by a perceived lack of justice over an initial offence, as well as strong in-group loyalties and, at least in the case below, competition over lucrative markets. As such they reinforce the need for strengthened law enforcement which can clearly demarcate resource entitlements as well as respond to crimes promptly. 5Of course, it is likely that many non-violent khalwat arrests went unreported. The ones which were reported this month involved either a beating or a bule (foreigner). 7 Box 3: Examples of vigilante attacks in November · "Moral vigilantes" and khalwat in Aceh Tamiang: Serambi reported that the Vice Head of the Aceh Tamiang DPRD was caught alone with a divorcée in a friend's rented house. Residents began beating the man but were stopped by the WH and police who soon arrived on the scene. The man was subsequently recalled from his DPRD position and ousted from his party (PAN). · "Moral vigilantes" and improper Islam in Aceh Besar: A mob from 18 villages descended on a pesantren (Islamic boarding school), vandalizing property and vehicles. Serambi reported that the reason for the attack was that the pesantren taught a "different" form of Islam, in which there was no Friday prayers, boys and girls read the Koran together from midnight to 4 a.m., and non-members were not permitted to observe. The head of the pesantren subsequently promised to move the school elsewhere. · "Retribution" vigilante attack in Lhokseumawe: A group of minibus drivers conducted "sweeping" of intercity buses, forcing them all to stop, after a minibus driver was beaten by a group of intercity bus drivers over a traffic incident. The minibus drivers also warned all intercity buses not to accept passengers for journeys within the city. 8