Despite the ubiquitous presence of militias in democratic politics, their influence on local elections in Indonesia is moderated by competing logics.
Indonesia’s New Order regime relied extensively on militias to control dissent and fight insurgents. Following Suharto’s fall from power, these groups have emerged as prominent actors in local politics.
Today, militias’ social, political and security functions defy easy classification. Some groups self-identify as ormas (community organisations) and maintain a national presence oriented around nationalist or religious ideologies. Others, known as pamswakarsa, are more local in nature and focus explicitly on maintaining social order through vigilante action. Yet others, such as the various “youth-groups”, are identified with urban racketeering and are often known as preman. Despite this variation, militias’ strong coercive presence and proximity to state actors endows them with tremendous “street-power”.
The possible deployment of militias’ street power for political ends was a chief concern during the last round of pilkada (direct local-executive elections), held on 9 December 2015, and simultaneously across 254 districts and nine provinces for the first time. The logistical challenges involved fanned fears of skirmishes over administrative irregularities. Moreover, ongoing internal conflicts within several political parties increased the risk of disputes over the nomination process.
Local police chiefs anticipated militia involvement in security disturbances and held regular meetings with their leaders. At the national level, ministers and top bureaucrats urged militia organisations to help “secure” the elections. Militias were quick to respond with solemn declarations of support for peaceful elections but continued to be a highly visible part of the electoral process. Some groups publically endorsed specific candidates and openly participated in campaign activities. Others claimed political neutrality and instead, vowed to deploy their ‘troops’ to assist the police and military in guarding the elections.
To understand the factors that drive militias’ influence on local elections in Indonesia, we observed their activities in four diverse districts. We found that despite their ubiquitous presence, militia’s involvement in elections is contingent on two factors. First, militias are more likely to display explicit political support for particular candidates in regions where they face competition from rival organisations over state patronage.
Second, militias are likely to engage in concrete action to support their political allies when the level of electoral competition is also high. From the candidates’ perspective, the tangible value of electoral services provided by militia organisations is quite limited. However, politicians continue to seek militia support as an “insurance” policy during the elections and beyond.
We followed campaign activities leading up to the pilkada in four districts: Medan, Manado, Surabaya and Central Lombok. This selection was made based on two criteria: level of inter-militia rivalry and competitiveness of the electoral race.
There is a dominant militia in two of the districts: Nahdatul-Ulama’s Banser in Surabaya and Brigade Manguni Indonesia (BMI) in Manado. The other two districts are, more fragmented. In Medan, there is intense competition among several “youth organisations”, most notably Pemuda Pancasila (PP) and Ikatan Pemuda Karya (IPK). Central Lombok is also home to several rival militias such as Laskar Ababil, Hizbollah, Amphibi and Buru Jejak Kumpul (BJK).
The level of electoral competition was high in Central Lombok and Manado where the races were contested by five and four pairs of candidates, respectively. Incumbents, who had support from broad party coalitions and did not face significant challengers, dominated in Medan and Surabaya.
Fluid political alliances driven by competition for patronage
Despite the New Order lineage of many militias, their orientation in democratic politics is not predicated on old loyalties. Instead, alliances forged by militias during local elections are fluid, based on tactical expectations of patronage from individual candidates rather than strategic alignment with a particular political agenda. Specifically, we found that militias are more likely to signal explicit support for specific candidates in regions where inter-organisation competition over access to state resources is intense.
In Central Lombok, the five candidates were supported by different militia organisations. Most notably, the winning pair Suhaili-Bahri from Partai Keadilan Sejahtera (PKS) and Gerindra, was supported by Laskar Ababil, Banser and Amphibi. Hizbollah supported the candidate from Partai Kebangkitan Bangsa (PKB) and Partai Nasional Demokrat (NasDem). BJK, arguably one of the most powerful pamswakarsa in the district, fielded a candidate as an independent. In Medan, all rival militias, including PP and IPK declared their support for the incumbent Eldin-Akhyar, who eventually won against a single challenger by over two-thirds of the vote.
Leaders of militia groups cited expectation of material and non-material favours as factors that guided their choice of candidates for endorsement. Frequently mentioned rewards for aligning with a candidate included access to Bansos (social assistance funds), preferred status for bidding on infrastructure projects, and employment for rank-and-file members through assignment of parking and security contracts. Additionally, militias expected a demonstration of symbolic “closeness” from candidates.
One militia leader in Central Lombok noted: “If [our organisation] invites him [the candidate] no matter how busy he is, he will attend.” In Medan too, militia groups deliberately held their local chapter elections during the pilkada season so that the endorsed candidate can inaugurate them. Perceived access to the candidate helps militia leaders’ improve their standing with rank-and-file members.
When a militia is dominant in a region, its leaders find little value in explicitly supporting one candidate over another. In Surabaya, NU’s Banser did not endorse the popular incumbent Risma. Even in Manado, where the electoral competition was extremely high between four pairs of candidates, the BMI insisted on maintaining its neutrality. In both cases, militia leaders were confident that regardless of who is elected, their organisation would have sufficient access to the walikota (the local executive).
When considering how competition for patronage affects militia involvement in the pilkada, it must be recalled that the district executive is one of several avenues providing access to state resources. In all four cases, militia organisations maintained close ties with the bureaucracy, police and military officials. It was not uncommon to find posters of militia leaders in combat fatigues, giving a thumbs-up with newly appointed police or military officials.
These relationships not only enable militias to position themselves as local partners for funding programs managed by state bureaucracies but also subject them to some level of state control. Thus, while fiscal decentralisation may have increased the value of accessing the local executive’s office, these alternative channels reduce the stakes involved in backing a successful candidate.
Competitive disadvantage in electioneering
While high levels of competition between militia organisations in a region can signal political support for a candidate, it does not automatically translate into any electioneering action. Active support by militia groups is highly contingent on the candidates’ willingness to pay for specific services, which is more likely in tightly contested races.
Indeed, militias were quite active in providing logistical support to endorsed candidates in Central Lombok. Activities typically involved arranging pawai (processions) through neighborhoods, accompanying the candidates to their public and private events and handling the food and transport arrangements for attendees.
In sharp contrast, militias in Medan were mostly detached from the process. Militia leaders explained their involvement in campaigns as purely on a “by order” basis. In previous elections, candidates sought their assistance in gathering attendees for public rallies and production of campaign materials such as banners and posters. However, leaders complained that the leading candidate Eldin-Akhyar had been particularly “stingy” with disbursing funds for large-scale campaigning. As a result, their members have not had much to do in this cycle.
From the candidates’ perspective, the value of acquiring militia support in elections is quite limited. The negative, preman (gangster) image of militias puts them at a disadvantage for mobilising votes from the public. Over-enthusiastic militia members also create disciplinary problems for a candidate with election commission officials, as they frequently get into brawls with rival groups and damage opponents’ campaign materials. Campaign teams claimed to limit militia involvement to ceremonial activities to minimize these risks.
Candidates were equally unimpressed with the militias’ ability to mobilizse votes from their members, who tend to have diverse affiliations with political parties. In Medan, when militia leaders have nominated themselves for a DPR seat or executive campaign, they have been dealt with embarrassing defeats even in the areas under their ‘control’. Similarly, in Central Lombok Buru Jejak Kumpul nominated its leader as an independent candidate in this election, who managed to secure only four per cent of the popular vote.
Finally, campaign teams were quick to rule out the possibility of using militia networks for disbursing vote-buying funds. A staffer explained that “the ‘leakage’ would be just too high if we did it [disbursement] through them [militias].” To ensure that the funds reach potential voters, politicians tend to rely on personal or ethnic networks that may or may not include individuals affiliated with militia organizations.
Voter intimidation during tight races
Despite a number of procedural disputes in several districts, the December pilkada remained largely peaceful. Within our four cases, we did not observe the overt use of violence against voters. Election organisers partially attributed this to the fact that the simultaneous elections reduced militias’ ability to mobilise members from different districts for potential disturbances. However, militias’ unique coercive advantage suggested the possibility that they may be deployed for low-level intimidation, especially in during tight races.
Indeed, in Central Lombok, the local police reported a pattern of cattle theft by militias to influence voter choice in villages that had pledged support to a rival candidate. A local police chief noted: “the goal is to create insecurity. They intimidate [the residents] so that they switch their choice [of candidate]”.
Voter intimidation also seemed likely in Manado, where the election became especially contentious due to the last-minute disqualification of a leading candidate. There was also a potentially dangerous communal dimension to the conflict: the disqualified pair, Imba-Boby nominated by Golkar and PAN (Partai Amanat Nasional) claimed to represent the Muslim minority in Manado while the hegemonic BMI militia has a reputation for guarding Minahasa’s ‘Christian’ interests. However, even as the mainly Muslim crowd staged sustained protests in the city and the election was eventually postponed, the BMI and its affiliates remained largely unconcerned. BMI leaders explained that their organisation did not have a preference for a particular candidate in this election and they did not gain to stand anything from getting involved.
Intimidation did not seem to be a factor in Medan due to the incumbent’s significant lead over his rival. However, several sources reported systematic intimidation by militias during the 2010 pilkada that was tightly contested by 10 candidates. According to a local politician, militia groups set guard outside Chinese neighborhoods in the city and prevented residents from voting for their co-ethnic, Sofyan Tan, nominated by the PDI-P. Tan eventually lost the election to the incumbent during a run-off.
Based on these observations, it appears that militias’ coercive functions are activated when there is a confluence of three factors: inter-organisation competition is high, the electoral race is close, and there exists a readily identifiable group of voters whose choice could be decisive in determining the outcome.
‘Insurance’ value of militia support in local elections
Militia organisations and their members are a highly visible part of the electoral process. However, their tangible political influence, in terms of political endorsement, voter mobilisation and intimidation is contingent on the level of inter-militia competition for access to local resources and electoral competition. Even so, the kinds of services militias can provide during elections are quite limited. Militias are not able to marshal votes for candidates due to their negative public image. They can be deployed for voter intimidation under certain circumstances, but the reputation costs of seeking their coercive support are not negligible.
Given the limited incentives and scope for militia involvement in elections, why do candidates continue to court them? Politicians seek militia partners not so much for their services in an election but more for their “insurance” value. The primary purpose of having a militia organisation in one’s camp, or keeping it neutral, during an election is to prevent it from supporting a rival’s campaign. As one told us: “They cause trouble sometimes but we would still rather have them on our side rather than someone else’s [side].” In this sense, the importance of militias’ role in elections is defined not only in terms of what they do, but also what they don’t do.
Finally, electoral alliances with militias seem to prove more valuable once a candidate wins office. Several candidates explained how having a militia on one’s side can offer a district-head ‘protection’ from extortion by other politicians in the local legislature and bureaucrats. This leverage is especially critical in preventing and managing corruption investigations.
This study was conducted with support from the research team at the Center for the Study of Religion and Democracy (PUSAD): Dr. Samsu Rizal Panggabean, Ayu Mellisa, Husni Mubarok, Irsyad Rafshadi, and Siswo Mulyartono.
Sana Jaffrey is a PhD student in political science at the University of Chicago. She previously led the design and implementation of the National Violence Monitoring System (NVMS) database at the World Bank during 2008-2013.
Ihsan Ali-Fauzi is the founder and director of the Center for the Study of Religion and Democracy (PUSAD), Paramadina Foundation, and a lecturer at the Paramadina Graduate School, Jakarta, Indonesia. His publications include “Learning from each other: Muslim Societies in Indonesia and South Asia” (2009), “Disputed Churches in Jakarta” (2011), and “Policing Religious Conflicts in Indonesia” (2015).