Printer Friendly

Local and national: keroyokan mobbing in Indonesia.

From horrific accounts of men decapitated to "ordinary" accounts of stolen motorcycles, the routine beating and killing of alleged criminals by mobs (massa) has become common in Indonesia. This article examines the patterns of keroyokan--mobbing--from 1995 through 2004 in four provinces and highlights the temporal, spatial, and substantive variations of this phenomenon. Drawing from a database of provincial and local news clippings in Bali, Bengkulu, West Java, and South Kalimantan and in-depth case studies and interviews, this article shows that mobbing varies considerably. Its causes are national and local. The temporal data show that nationally the most important factor to influence levels of mobbing was the introduction of decentralization. The power vacuum that resulted from the policy decision to transfer authority from the center to localities increased local violence. Yet this macrolevel explanation is inadequate to show the spatial variation and different forms of mobbing violence. To understand the causes of these dimensions of variation, one has to move away from macronational approaches measuring violence and include a more microethnographic local approach. A richer understanding of mobbing must be locally rooted. This article uses three case studies to illustrate the centrality of local factors affecting this form of violence. The case studies suggest that mobbing is shaped by the acquiescence of actors in local communities and local learning. The article draws attention to the need to incorporate local data and methods into an analysis of violence in Indonesia and to appreciate varied daily rituals of violence as reservoirs of conflict.

KEYWORDS: Indonesia, keroyokan, mobbing, violence, Bali, Kalimantan, Bengkulu, Java, ethnography, decentralization


The prominent Indonesian writer Goenawan Mohamad highlights an "ordinary" dimension of violence in contemporary Indonesia.

Each murder shines a spotlight, and thousands of deaths are going on in the shadows. The mass media informs us of them, but the mass media also distances them. On such and such page in such and such a column of this or that newspaper, at this time in that program on television, an event, something of great profundity, has turned into a fact, something level, flat.

And even savagery becomes level and flat, one dot in one row, a dot the same as all the other dots.... And suddenly we are aware that cruelty is not something distinct. No, we are not the wild, wild west. In the world recreated in John Wayne and Clint Eastwood films, death is a possibility for anyone, for no one monopolizes violence, in our world today anyone, casually, is a potential victim, and anyone ... is a potential executioner. Cruelty crouches outside the door, ready to knock, asking to come in. (Goenawan 2002, 235-236)

Unfortunately, democratization did not bring about lower levels of violence; newspaper accounts describe incident after incident of killings, whether they are in traumatized Aceh or on the streets of the capital, Jakarta. Even now in 2008, after two competitive elections and the expansion of the rule of law, local mobbing persists. It is often missed in the study of violence in Indonesia.

As other articles in this issue show, scholarly attention has centered on the conflict areas--Maluku, Aceh, Papua, and Central Kalimantan (1)--or on particular types of violence such as jihadi (religious vigilantes) or terrorism. Violence in ordinary communities has largely been ignored. In this article, I aim to refocus attention away from the headlines of conflict to the regular inside columns, or "dots" as Goenawan's label suggests. The dot in question is keroyokan, or local mob violence.

Understanding keroyokan involves studying this form of violence at the national as well as local levels. Drawing from a dataset based on material collected in four provinces, I trace the relationship between keroyokan and national political developments since 1995. The evidence suggests that the main factor contributing to increased levels of violence was the implementation of the decentralization policy of 1999. When the policy took effect, there were sharp spikes in local mobbing, due in part to the power vacuum in law enforcement/local bureaucratic institutions as well as the opportunity to carry out revenge (dendam). Changes in the pattern of law enforcement, patronage networks, and conflict resolution mechanisms were all transformed with decentralization and created conditions for increased levels of mobbing. Yet the national context does not account for the spatial distribution and different forms of local mobbing; it fails to completely capture the variation in mobbing. Local conditions provide a richer answer: the tacit, and sometimes explicit, approval of local elites of the use of mobbing, the acquiescence of local law enforcement to mobbing, and learned behavior of mob violence as conflict resolution (stemming from repeated patterns of violence in specific communities) affect the pattern of mobbing. Using a combination of national and local factors, I provide insight into the persistence of ordinary violence across Indonesia. Understanding violence at the local level is especially important, because the predominant pattern of violence post-1998 is changing from vertical state-induced to horizontal society-society violence (see Tadjoeddin 2000; van Klinken 2007).

My approach differs sharply from that of the other articles in this issue, in that I combine macro- and microlevel analyses of conflict. I show that in order to see and understand the dots, one has to adopt a more effective methodological lens, an ethnographic approach. This approach has been commonly used to study violence elsewhere but has little application in Indonesia. (2) As such, my study combines a macrolevel quantitative approach with a rich microqualitative analysis, using both a dataset gathered from local newspapers and intensive fieldwork and interviews. It shows that an ethnographic approach is effective in illustrating how a conflict evolves and the critical catalysts that shape the form and level of violence. I begin the article with a discussion of mobbing and the methodology adopted to study this phenomenon. I then briefly describe the temporal, spatial, and substantive variation in mobbing in Indonesia and lay out the three local case studies of keroyokan. Turning to the causes of the phenomenon, I conclude by laying out the national and local factors at play, highlighting the needs for a synergy of approaches for an in-depth understanding of why and how local mobbing occurs.

Introducing Keroyokan

Most Indonesians are familiar with keroyokan killings--a ritualized form of violence that involves the practice of an unlawful group, usually ordinary citizens, controlling and punishing crime usually in the form of a lynching or a severe beating. Newspapers reports, especially those in the tabloids, feature keroyokan regularly, and contemporary crime television shows often highlight cases of citizens taking the law into their own hands through "mob justice." This "violent ritual," to use Charles Tilly's (2003) term, involves three components: a mob (massa); a violent entrepreneur (or more than one)--the mob leader(s)--who instigates and inflicts harm on the victim(s); and the victim(s), usually male, who may or may not be guilty of the alleged crime.

The ritual follows a similar pattern: a small mob composed of men and women captures the victim (usually a male between the ages of 20 and 35), beats the victim to death or seriously injures him (sometimes after a shaming ritual), and then, in some cases, sets him on fire. Incidents include the beating of local thieves for simple things like stealing a shirt or a swallow's egg to more serious crimes involving rape and witchcraft. The practice of keroyokan--literally in Javanese "the ripping apart of people"--goes beyond the narrower Western interpretation of the terms vigilantism and lynchings.

Keroyokan should he understood to be an umbrella term that covers a diverse set of social phenomena and types of violence. Although the ritual is similar in that the three core elements--the mob, the violent entrepreneur(s), and the victim(s)--remain consistent, differences exist in the catalysts, social networks, and coordination associated with different incidents; the levels of brutality; the targets; and the social legitimation within communities where the incidents occur. It is important to understand how mobbing varies in form to fully appreciate why it occurs in specific forms in some locations and not others. The variation in substance laid out typologically below is drawn from my analysis of over 4,000 cases of mobbing.

The catalyst difference is straightforward--mobbing is motivated for different reasons. Mobbing can be spurred by incidents involving theft, by more serious robbery/extortion, by hit-and-run accidents, and by rape--all under the rubric of crime. A second rubric involves interpersonal relations that are provoked by acrimonious exchanges between individuals in a local community. Within this rubric are cases of intergang warfare, ninja killings (involving targeting of specific individuals for assault/assassination, usually involving the perpetrator disguising his or her identity), and revenge for personal harm--that is, alleged adultery, shaming, or conflict over property. A third rubric involves violation of societal norms. Victims are targeted if they are seen not to conform to the norms of local behavior, as manifested, for example, by alcoholism, mental illness, mental handicap, the practice of witchcraft, or other violation of local custom. Crime, personal acrimony, and norm violation broadly capture the range of catalysts of mobbing.

A second substantive difference is the level of premeditation. This captures the depth of social networks and level of coordination that are involved in carrying out the ritual. The image of "running amok" has falsely created the impression that mobbing is always spontaneous. (3) Many cases of keroyokan, especially those involving established personal relationships and norm violation, are systematically planned. The timing of the event, the rumors that are placed to discredit the victim, and the coordination among the violent entrepreneurs require systematized organization and coordination. To capture this variation, I distinguish between three levels of premeditation (low, medium, and high), which ranges from spontaneous responses to extensive organization over days, sometimes months. Incidents involving witchcraft, for example, are systematically premeditated, while cases of hit-and-run accidents are more reactive, more spontaneous responses of witnesses to the accident.

Equally significant is the level of brutality. On the surface it may be difficult to distinguish between physically beating a person and burning the helpless victim alive. Here, too, there are subtle, but important, distinctions. In cases classified as "low" or "medium," such as petty theft and hit-and-run accidents, the aim is to punish, not necessarily to kill. This is not to say that killing does not take place in these types of cases, but rather that the intentionality of serious injury is not as defined in advance. This pattern is most evident in adultery cases, where the practice of keroyokan can involve a shaming ritual. Rarely does keroyokan lead to death in these circumstances. This is differentiated from "high," where participants aim to kill from the onset. Two outliers, labeled "very high," that reach even greater levels of pain infliction are found in witchcraft and ninja cases. In these types of cases, the level of torture and physical bloodshed is considerable. In a Bogor morgue in August 2002, the level of pain inflicted on a victim was seen to be intensive; the decapitated body had the legs chopped off and no less than 100 slashes. (4) The autopsy revealed that most of the slashes had occurred while the victim was alive. The head was burned, but the skull showed that it had been cracked open earlier while the victim was conscious. The autopsy report estimated that the victim had been tortured for over three hours before he died. Ninja killings follow a similar pattern. What appears to be occurring in these cases is the use of increased brutality to exorcise spirits from these witches or ninjas.

A fourth substantive difference in keroyokan is the target, the victim. In almost all the forms of mobbing the victims are male, usually between the ages of twenty and thirty-five. The only cases involving women are witchcraft cases, in which the victims are usually older. In most forms of mobbing, the victim is known to the mob, an insider. Yet the high cases of petty theft usually target outsiders, not known to the mob (except in rural areas). Hit-and-run accident victims are almost always outsiders.

The social legitimation of the different forms of keroyokan also varies and has important implications for repeated patterns of violence in a community. Most incidents involving norm violation (except local custom) and personal acrimony often trigger a vicious cycle of violence in a community. They are seen in the community as murder rather than community justice. In these cases, there are strong lingering perceptions that a moral code was violated and the incident of mobbing was not justified. In Tanggerang near Jakarta, for example, regular patterns of gang violence over territory spill over into the community, with shopkeepers and stall owners forced to choose sides and often inadvertently participate in the violence. (5) Incidents involving local custom and crime are deemed to be the decision of the community as a whole and often do not provoke responses from the deceased victim's family or social networks. This form of social legitimation deems the actions of keroyokan "acceptable" within the norms of the community. The acceptance or rejection of mobbing in a community leaves an imprint for possible future violence.

These differences are laid out in Table 1, which shows that all keroyokan is not the same in substance. The form of violence varies considerably. It is important to lay out these differences in order to understand why mobbing occurs in some areas and not others, and to understand the implications on future violence in communities.

Understanding Keroyokan: Data and Research Methods

This article is part of a study that focuses on four questions: How often is keroyokan occurring, what forms does the violence take; why is it occurring; and, especially, what can be done to prevent it? The research is ongoing, and this article represents a preliminary analysis of findings. The final project will include an assessment of areas within the provinces studied where keroyokan has not emerged, and show why these areas have escaped this type of violence. This article discusses the first three questions--the frequency, forms, and causes of keroyokan violence--drawing from the data in the four provinces and case studies.

Drawn to the subject of vigilantism by the accidental witnessing of a decapitation in a coastal town in Banten in 1999 (Welsh 2001, 2003), I chose as my first challenge to compile a dataset that would capture the frequency of mobbing nationally. This challenge was especially difficult, since many cases of mobbing often go unreported, especially in rural areas. Although the research focus was on the period of the democratic transition (post-1998), I began the analysis in 1995 in order to examine how levels of mobbing corresponded to Indonesia's democratization process. Faced with the challenge of poor records, confidentiality questions, and conflicting "truths" in recounting events, I compiled three different datasets. (6) The most complete (which forms the basis of the macroanalysis) was gleaned from local newspaper accounts. (7) Four provinces were chosen to capture national patterns. They were specifically selected for regional variation and in nonconflict parts of the country. The provinces, which represent each of the major regional zones of Indonesia, are the populous western islands of Java and Sumatra, the northern island of Borneo, and the eastern island of Bali. Within these zones, the provinces selected were reported to have some of the lowest levels of violence--that is, they were not hot spots of ethnic or religious conflict. The plethora of newspapers that emerged after 1998 facilitated the accounting process. The focus was on local crime reporting, especially in the tabloids, since this medium provided the most complete, albeit sensationalized, local coverage. (8) Because many of these newspapers fail to keep copies of issues past a few years and most are not computerized, these records have gaps that go well beyond the oversights of painstaking scans over daily issues. As such, the numbers I report here should be interpreted as indicators of trends rather than completely accurate statistics.

The dataset providing an indication of macrotrends was complemented by selected microlevel case studies and interviews focusing on specific incidents. Incidents were selected based on variation, and the areas were chosen based on the frequency of keroyokan cases. The aim was to compare the newspaper reports with accounts of at least two of the actors associated with an incident--the police officer, village leaders, mob participants, and family members of the victim. These local interviews provided insight into the different causes of mobbing in particular incidents. In total, thirty case studies were selected, with a minimum of six in each province. Here, too, there were shortcomings, especially with regard to cases during the New Order period in which witnesses often did not recall the exact details of the incident or constructed a new version to explain the event, which often did not mesh with the specific details in other accounts. Parallel "truths" of accounts coexist. Due to the time involved, most of the police officers were transferred outside of the area, requiring interviews in alternative locations. More often than not, however, their accounts were not available. Case studies in areas with high levels of migration, urban traffic, or transitory populations were also difficult to assess accurately. As memories faded with time, accounts varied even more widely over five years. Not surprisingly, most of the complete case studies are from 2000 onward or involve "well-known" cases in specific areas with relatively stationary populations. When possible, the local interviews were accompanied by interviews with journalists, local and national officials, nongovernmental organization (NGO) activists, and social leaders, all with the aim of providing a rich combined set of macro- and microdata to examine and compare national patterns of keroyokan. Three of the case studies are summarized here.

Patterns of Keroyokan: Temporal, Spatial, and Form Variation

Local newspapers in the four provinces paint a disturbing picture of the frequency of mob killings. Newspaper accounts, detailed in Table 2, record 4,037 incidents involving 5,506 victims, from 1995 through 2004. Nearly a quarter of the incidents (23.7 percent) involved more than one victim. In one instance, in southern West Java in 2000, twenty-three people were killed in an afternoon. (9) The data are from a small cross section of Indonesia, so the total national figure is much higher. This is particularly significant since the four provinces were selected because they were seen as less violent than others, such as Central Kalimantan and East Java. Other research suggests similar, even higher, levels of mobbing in other parts of the country (Madden and Barron 2002).

Figure 1 traces the number of incidents and victims in the four provinces over time and shows that the combined number of incidents/victims increased sharply after 1999 and has decreased since then. Yet, both the numbers of incidents and victims have increased since 1999, compared with the New Order period. The year 2001 is important as a pivotal turning point nationally in political conditions, the impact of decentralization.


Table 3 and Figure 2 outline the pattern of keroyokan in the four provinces studied. They show that the levels of keroyokan vary considerably over the four provinces. West Java, with a higher population, not surprisingly has the highest number of cases, while Bali reports the highest number of incidents per capita. South Kalimantan and Bengkulu report lower frequency of incidents, with the smallest province of Bengkulu reporting the fewest. The pattern of the frequency of mobbing in three of the studied provinces--West Java, South Kalimantan, and Bengkulu, which reported the highest frequency of cases in 2001--corresponds to the pattern of the four provinces as a whole. In South Kalimantan and Bengkulu, the increase is less sharp than in West Java. Bali does not follow the broader pattern and reported the most cases in 1999. The spatial variation of mobbing highlights the need for incorporating national and local assessments to determine its causes.


To illustrate the spatial variation within provinces, the incidents of mobbing from 1995 through 2004 in Bali are mapped below (Figure 3), weighted for district population. The eastern district of Bali, Karangasem, has the largest concentration of incidents, while cases in western Jembrana are rare. Both districts, which are rural and are a considerable distance from the economic and political center of the island in Denpasar, have extremely different patterns of local mobbing. It is this regional spatial variation that points to the need to examine local conditions in order to understand why mobbing occurs in some areas and not others.

Coupled with the spatial variation are differences in the form of keroyokan found in specific localities. The macrodataset shows that the majority of incidents involve crime: petty theft (48 percent); hit-and-run accidents (0.5 percent); robbery, extortion, or drug pushing (21 percent); and rape (3 percent). Yet, many of the catalysts are acrimonious interpersonal conflicts, such as revenge over alleged adultery (1.5 percent) and shaming or conflict over land or water rights (5 percent). The personal revenge motive, dendam, lies at the heart of many incidents, especially in rural areas. (10) Others involve the behavior of an individual who is deemed not to conform to local norms and who has been accused of inflicting harm on others. Such a person could be a drunkard, be mentally ill or mentally handicapped (2.5 percent), or be an inept practitioner of witchcraft. "Witch doctor" or dukun santet (6 percent) victims have allegedly harmed others in the community. (11) Finally, some cases are tied to local custom and are carried out on the basis of a community decision to punish for violating custom. These cases are found in specific regions, such as eastern Bali or South Kalimantan. The variation by catalyst is shown in Table 4.


The data show that there is a connection between the concentration of levels of mobbing and the forms of mobbing. For example, in Karangasem in northeastern Bali, there are four villages that have high concentrations of keroyokan over the ten-year period 1995-2004. In total, 122 people were killed through keroyokan in eighty-four incidents. The overwhelming majority of these cases (94 percent) involved local custom and contestation between villages over water rights, leading to a cycle of revenge between villages. Interviews in these villages reveal that leaders did not oppose the practice of keroyokan, and in fact encouraged the practice. (12) In one case, one of the village priests, a local elder, was the violent entrepreneur. He was the key individual egging on a mob in at least seven cases.

In a parallel example in Indramayu in Western Java, there are ten villages with high concentrations of keroyokan. In the ten-year period, there were eighty-four cases of mobbing in which 103 people were killed. The overwhelming majority of these cases involved petty crime and gang competition between villages. Here, too, interviews highlight the critical approval of local leaders in the practice of mobbing. (13) Local village leaders described this practice as "normal" and appropriate for addressing social tensions. In one interview a local leader claimed that mobbing was a "necessary pastime for young people to prove their value in the local community," almost a rite of passage for social acceptance. While few local leaders actually participated in mobbing, in contrast to the villages in Karangasem, participants in the mobs interviewed reported that they were praised by village leaders for protecting their community and were given social status for their actions. The organization of these villages as "gangs" evolved over time and reflected growing competition between villages for access to economic wealth--namely, the export of labor to Malaysia and involvement in the prostitution trade. Rising wealth and competition in this area fueled a cycle of mobbing approved by local leadership.

Keroyokan Under the Microscope

The substantive differences in keroyokan are perhaps best illuminated by drawing from the ethnographic case studies. I briefly describe here three incidents, drawing from interviews in these localities, each under a different rubric: personal acrimony, norm violation, and crime.

Case 1: Revenge for Adultery

In the village of Gambut outside Banjarmasin in South Kalimantan, Lidya and Hip were caught by residents in a compromising situation, coming out of a rented room in the housing district of Pankaria on December 29, 1998. (14) Residents had been watching the couple for weeks. They planned to punish Hip for his violation of local norms.

Lidya was married to a local businessman, an older man respected in the community. She had married her husband, Jako, when she was seventeen years old in an arranged marriage organized by her parents, who were also respected members in the community. Lidya's beauty was well known, and Jako had waited for the earliest opportunity to marry her, offering bride price to secure her parents' cooperation. Jako had by all accounts treated his young wife with devotion, showering her with gifts. His first wife had died in childbirth at the age of twenty-five, and he had waited over twenty years to marry again. Jako's family described at length his longing for Lidya and his desire to have a family with whom to share his success. He had apparently watched Lidya grow up and bloom into a beautiful young woman, and when the moment was right, in mid-1997, he asked for her hand in marriage.

The first few months of marriage were reportedly blissful, according to Jako's sister. But by the end of that year, with no pregnancy on the horizon, tensions were developing in the household. It was at this point, in January 1998, that the young handsome Hip came on the scene. He was an outsider, born in Central Kalimantan, yet a distant relative to Jako's neighbor, Atma. The neighbor agreed to take in the twenty-two-year-old if he would help in the fields and around the house.

Lidya and Hip soon met and fell in love. Hip reports that he fell in love with Lidya the first moment he saw her but held off showing his feelings. He claimed that it was Lidya, in a loveless marriage, who made the first move. The neighbors and Jako's family, however, place the blame on Hip, the outsider, who in their view corrupted the "innocent" young woman. It turned out that Hip was actually married, with a wife back in Central Kalimantan.

Jako was apparently unaware of the budding relationship in the first half of 1998. It was not until he went on a weeklong business trip in July that the affair began. Discreetly at first, Hip and Lidya began seeing each other. Hip "borrowed" a room from a trusted friend for the affair about half an hour from their residences. For six months, the affair deepened, during which Jako's business travel became more frequent.

Neighbors soon noticed Lidya's absence from home. One morning in early December, Jako's nephew Mat followed her and discovered the lovers' hiding place. He shared the news with his mother, Jako's sister, and a small group of "concerned" neighbors. For three weeks, neighbors and family members discussed what action to take, what message they could send to express their anger and disdain. The focus was on punishing the man, the outsider. Mat consulted with family, neighbors, and even local police officials on options. He reported that he told Lidya's parents of the affair and pledged that he would punish Hip appropriately. Her parents gave tacit approval. Mat also reportedly met with the local policeman, essentially asking for permission to handle the situation on his own. Mat reports that the officer not only approved of the need for the family to take action first but also pledged his assistance in prosecuting Hip subsequently. He met with the local religious leader and village headman as well, who both called for punishment. Mat decided, however, not to tell Jako of the affair, fearing that the pain of discovery would be too much for him and possibly endanger Lidya. The community would act for him, Mat reported.

They bided their time, as three neighbors took turns watching Lidya. It was the holiday season in this mostly Christian area. The moment of reckoning came on December 29, in the early afternoon. Lidya was followed into the Pankaria housing district. Mat reports that he was not the person who saw her in the room but knew "at once" that the time was right and that she was with Hip. He called the neighbors and his friends, amassing a crowd of over twenty people. Armed with sticks they shouted and surrounded the house, calling out to Hip, who appeared at the doorway dressed humbly in a sarong. They dragged him out of the house and systematically beat him. Lidya screamed but was rushed away by Jako's sister and another female neighbor. The neighbors shamed and publicly humiliated Hip for violating his marital oath and the marriage of another man by parading him through the streets. The neighbors called out to Atma, Hip's relative, who was asked to pledge that Hip was no longer welcome in the village; Hip was then turned over to the local police. He suffered two broken ribs and countless other minor injuries.

Both Lidya and Hip were taken into custody for questioning, although Lidya was soon released. She and Jako moved out of the area, and Hip moved to the Pankaria district, where he still resides. Mat became the caretaker of Jako's local business. Hip was released without charges, and his physical injuries healed.

Case 2: Witchcraft

Nearly two years later, on August 24, 2001, in the village of Cimanggu Dua south of Bogor in West Java, a forty-five-year-old man known as Saprudin was killed after he was accused of dukun santet, or black magic. (15) This incident has a level of premeditation similar to that of the adultery case but was much darker.

Saprudin was born in Cimanggu Dua to a modest household of rice farmers. He was reportedly not a good student and regularly got into trouble in his early twenties. He settled down, however, when he met Siti, a woman who was raised in the city of Bogor. They married and Siti moved into the village. She was believed to harness mystical powers and began to teach the young Saprudin witchcraft. In 1998, Siti died in childbirth delivering a son who also died. The neighbors believe that when she died, she passed on her powers to her husband.

Saprudin turned more to witchcraft in the years after her death, converting his talents into a profitable operation. He began to take on more requests, from removing spirits to inspiring jealousy and sickness. His reputation as a dukun santet grew exponentially. One event in particular, in early 2001, reportedly served to reinforce his power. Allegedly paid by a jealous husband, Saprudin cast a spell on the wife, who apparently turned overnight into a wife with unquestioned devotion to her husband. Saprudin's success served to reinforce his zeal for witchcraft, and he began to engage more in animal sacrifice and other forms of black magic.

In early August 2001, his neighbor Hussin complained about the stench coming from Saprudin's house, apparently from the dead carcasses hung in his yard. Hussin described Saprudin with a sense of loathing, mixed with fear. Acting as the "violent entrepreneur," Hussin met with the local religious leaders at the mosque. He sought out their advice on how to address Saprudin's growing power; Hussin's own small farm was negatively affected by the environment created by Saprudin's witchcraft.

Over the course of three weeks, Hussin brought key leaders from the village to see firsthand Saprudin's residence and observe him from afar. The process of building social legitimation was a critical part of the planned mobbing; it was necessary to build collective support to carry out the ritual against the norm violator. Hussin's wife, Sharifah, reported her fear of Saprudin, who in her view was able to bring the carcasses back to life, and whose eyes had turned from brown to green. Apparently rumors of Saprudin's "evilness" intensified after he reportedly confiscated a soccer ball that came into his yard, scaring the children playing in the field near his house. The neighbors in the area palpably reported their fear of Saprudin's powers.

For three weeks, the momentum to remove Saprudin built, as religious and community leaders were brought on board. On the morning of August 24, twelve men entered Saprudin's house after their prayers but before the victim regularly woke up. Taking care not to look into his eyes, they blindfolded him and dragged him out of the house. Outside a crowd gathered, armed with parangs (long knives). They systematically slashed him, beating him repeatedly. They beheaded him in order to release the spirit of evil in him. Mob participants then played a game of soccer with his head, turning over the head to the local children to join in the "game."

In recollecting the incident, Hussin describes the community's acceptance of the removal of the local witch, whose practice in black magic violated norms. He recalls newspaper accounts of similar incidents in nearby areas, couching the mobbing incident in the context of enforcing moral purity called for by the local Islamic leaders in the community.

Case 3: Petty Theft

In the sleepy area of Indramayu in northern West Java, the call for removing threats to the community took on another form. (16) On March 18, 2003, Shaiful was burned to death in a mob led by his close childhood friend Osman. Shaiful was unemployed at the time of his death at the age of twenty-one. He had searched for work in the neighboring city of Cirebon but had not been able to secure a job. He had returned to his village in August 2002, where he lived with his sister Mariah.

Shaiful began to engage in a series of small robberies in the community. It is not consistent in the accounts whether he had a drug problem, but it does appear that the petty theft was a regular occurrence. He stole a radio, a cell phone, and a wallet. Mariah cried in her recollection of the events as she laid out her frustration with her brother, who would not stop stealing. She talked about how she lost face in the community and how her family's reputation was harmed. She described her helplessness in the first few months of Shaiful's continued petty theft.

Mariah sought out Osman, who grew up with Shaiful and was a schoolmate and neighbor of the family and a trusted confidant. Unlike his troubled friend, Osman had managed to secure a job in Cirebon. In January he returned to his home village and spoke at length with Shaiful, warning him that his actions would have to stop. Unfortunately, they didn't.

In mid-March, Mariah called Osman once again and asked for his help. She reported that Shaiful had tried to steal a neighbor's bicycle and Osman needed to intervene. Osman returned home again and consulted with family, the local headman, and local religious leaders. It was collectively decided that Shaiful needed to be punished for his actions, that his theft was doing too much harm to others. Osman described at length his recollection of newspaper accounts of local justice and in his interview categorically pointed out that mobbing through burning had become a legitimate means to address crime. In his words, "Justice needed to be served." When asked whether he opted for the state's judicial system and local law enforcement, he scoffed and ridiculed these institutional options.

He described the focused means by which he killed his childhood friend. On the evening of March 18, Osman and approximately five members of the community confronted Shaiful about his theft. His sister Mariah watched a few yards away. Shaiful confessed. Osman asked whether he was ready for his punishment and asked him to beg God's and his sister's forgiveness. Shaiful was petrified, recounts Mariah, and asked for forgiveness. At this juncture, Osman poured gasoline over Shaiful, who at this point was fighting to escape. He was held down by the mob. Osman lit a match and threw it on Shaiful. It went out before it hit his gasoline-soaked clothing. Three matches later, one caught. Osman, Mariah, and the crowd watched as Shaiful burned to death. With his burning, Shaiful would never enter heaven and would continually be punished for his crimes. Mariah cries when she tells this story but believes that this was the "right" thing to do. Osman reported that he has been part of more than five mobs and has played the violent entrepreneur role in three. His status as a moral adjudicator in the community rose.

Keroyokan and Causes of Violence

Each of these cases shares the pattern of the common ritual. However, the form of mobbing was very different. The adulterer was spared, but the thief and witch doctor were brutally killed. All three incidents involved planning and consultation with local leaders, who gave tacit approval of mobbing. Yet, the level of state involvement remained minimal, as these events were shaped primarily by individuals in society, and the catalysts were quite distinct.

The macrodata earlier pointed to the variation over time, space, and form of mobbing. The ethnographic case studies illustrate the variation in the form of mobbing in more detail and the personal nature of the killings. What then accounts for mobbing and the difference in levels, spatial distribution, and form? I argue that the answer lies in a combined national and local analysis.

Research on violence in Indonesia has pointed to four sets of theoretical explanations. (17) First, studies have pointed to the critical role of the state as perpetrators and catalysts for violence. Violence in the New Order period of Suharto (1965-1998) was seen to have its origins in the actions of the military, who have been blamed for carrying out violence. (18) The critical role of the state has extended to the democratization period as well, where the state moved away from being the agent of violence to being an arena controlled by agents in society through electoral competition and checked by accountable law enforcement. The arguments here are twofold. Democratization ushered in a period in which the state was contested over, and violence was an outgrowth of this competition--often among local leaders with ties to national actors (see Aragon 2001; Bertrand 2004; Emmerson 1999; O'Rourke 2002; Wilkinson 2004). In another vein, the transformation of the state from military control in local areas to civilian policing created a vacuum in law enforcement, a weakening of state capacity. (19) This change took place, it is argued, during the regime transition of 1998-1999.

A second set of policies focuses on state policies rather than on the presence of the state itself. Among these explanations are the impact of the economic crisis and the mismanagement of the economy, and the introduction of decentralization. The former explanation ties violence to economic hardship, reminiscent of Ted Gurr's (1968) relative deprivation thesis. Suharto and subsequent presidents of Indonesia, B. J. Habibie and Abdurrahman Wahid, failed to revive the economy effectively, fostering conditions for crime and responses for "justice" by local communities. (20) The second policy, of decentralization, points to the changing structure of institutions in the state, moving power from the center to the periphery. (21) Decentralization was seen to exacerbate the power vacuum noted earlier, in that it intensified the problems of state capacity and also provoked tensions in peripheral zones, such as Aceh and Papua, as part of a disintegration of the Indonesian state. Both the critical role of economic conditions and the changing political structure were seen to contribute to violence post-1998.

A third set of explanations focuses on societal actors and their interaction. Analysts describe the collective action and group dynamics that shape group relations. Collective action theorists point to the critical role of leaders within a group, network building, and common identity--usually ethnicity--in shaping violence. (22) The framing of events, shaped by media and historical learning in a community, and the influence of individual violent entrepreneurs as the major conduits that interpret the need for the use of force, lead to violence. (23) Another parallel interpretation points to a common "violent" political culture, pinpointed as one that is willing to accept violence, as the factor underlying violence (see Colombijn 2002). In another vein, scholars point to the critical role of social capital--horizontal ties within a community--as an explanation for violence (see Varshney 2003). The focus is on the relationship within a community as either contributing to or causing violence. These explanations share a similar emphasis on societal relationships, although they differ in their emphasis on why collective action leads to violence.

A fourth set of explanations focuses on individual behavior. Psychological factors might lead an individual defined in the context of mobbing as the "criminal type" to instigate and participate in mobbing, sometimes as a result of trauma. (24) The idea of an individual running "amok" due to personal motivations is similar. (25) Rational self-interest is seen as another motivation: the use of mobbing can evoke personal satisfaction in the form of revenge or can yield economic or, in some cases, social rewards. (26)

Explanations that center on state and government policies often are national in scope, while explanations that are societal can combine local and national causes. Individual-oriented explanations are inherently local, although they can occur in a variety of localities.

Local and National Synergy: Keroyokan Explained

How then does the breadth of explanations of violence in Indonesia relate to keroyokan? The discussion earlier pointed to three major points of variation: temporal, spatial, and form. The macrodataset revealed that mobbing increased from 1997 through 2001 and then decreased. This pattern occurred in all of the provinces except Bali, which experienced the high point in violence in 1999. One of the major explanations of the variation in mobbing lies with what occurred in the three provinces in 2001, but earlier in Bali. A second important area of variation is spatial: Why is violence more intense in some areas than in others? What accounts, for example, for higher levels of violence in the district of Karangasem in Bali than in Jembrana? A third important area of difference is the form of violence. Why does mobbing take on certain concentrated forms in some areas and not others?

I turn now to each of these three points of difference individually. Rather than cover the gamut of explanations of violence, I focus on isolating some of the consistent factors that stand out in explaining the variation.

The temporal difference in levels of mobbing appears to be tied most closely with the implementation of decentralization policy. (27) B. J. Habibie introduced this policy through Law No. 22 of 1999. The new pattern of regional autonomy set in place fundamental institutional transformations for local bureaucracies and gradually increased fiscal autonomy. This was coupled with measures that changed the military territorial command's structure to increased civilian policing. These institutional changes took place while the political role of the bureaucracy was transformed from one loyal to Golkar, the dominant party during Suharto's New Order, to varied political allegiances, both triggered by new laws in 1999. Political pluralism emerged within the bureaucracy as part of the gradual democratization process. The institutional impact of decentralization was profound, in that the scope of authority and the roles and patterns of governance were transformed. The change, however, did have an impact by fostering a power vacuum in three areas: law enforcement, patronage, and conflict resolution. Community policing, often resisted by the military, evolved gradually to dismantle the territorial command structure. Patronage patterns were reconfigured as part of the electoral process, nationally driven initially in contests for parliament and the president and with the expansion of local gubernatorial, district, and mayoral elections, with greater depth in localities. The transformation in governance is evolving, in that the issues of corruption persist, yet most analysts agree that increased accountability has raised public awareness of this concern.

From the perspective of violence, and mobbing in particular, the institutional transformation has three major effects in localities. First, the deficit in law enforcement encouraged many local communities to impose their own forms of "community justice." Even today, the number of police per capita in Indonesia remains extremely low, with a ratio of over 50,000 per policeman. Second, the changes in patronage fostered conflict, which emerged from greater local competition, but simultaneously created uncertainty locally as to who would emerge as leaders in communities. Third, the transfer of authority from one group to another removed individuals as arbitrators of conflict, weakening local conflict resolution mechanisms. Since the New Order government had hardened who held power in communities, through ties to the military and Golkar, the transformation was significant in that new space had emerged for new leaders to emerge as previous leaders were discredited.

The comparison of the experience of Bali with the other three provinces suggests that decentralization did have an important impact on levels of violence. (28) Bali was one of the first places where the decentralization policy took effect. In late 1999, riots exploded in the main city of Denpasar, resulting from contestation over political patronage between Golkar and supporters of Megawati Sukarnoputri's party, PDIP. The trigger was the decentralization policy a few months earlier, which allowed bureaucrats within the state in Bali to oppose local Golkar members, many of whom had strong ties to Java and Jakarta. Law enforcement officials in Bali point to their relative incapacity to address the violence, since the military territorial command structure was in place and the police reform process was at a nascent stage. It would not be until after the 2002 terrorist bombing in Bali that policing would become more professional and reforms would deepen. The riots were concentrated in the island's main city, which comprises the most outsiders from Java and where the local conflict mechanisms were most tied to the national state rather than to local religious leaders. It is in this context, the early decentralization effect, that levels of mobbing spiked. The same transformations would not take effect until much later in the other provinces, due to less political competition, fewer potential gains from patronage, and greater military resistance to reform. Nationally, the policy of decentralization had the deepest impact on levels of mobbing. As policing and political pluralism took effect, creating more centers for conflict resolution, levels of mobbing dropped.

The national effect of decentralization, however, cannot account for the spatial and form differences in keroyokan. Here, local factors emerge as important. When one draws from the case studies and spatial patterns of concentrated areas of mobbing, two factors stand out: the acquiescence of actors in local communities, and learning of mobbing in a community.

In each of the three case studies examined, the violent entrepreneur sought out leaders in the community to gain tacit approval of the use of mobbing. This was true of Mat in the case of adultery, Hussin in the killing of Saprudin for witchcraft, and both Mariah and Osman in the killing of Shaiful for petty theft. They acted with the approval of leaders. Who they consulted differed, yet they included a broad range of society--religious leaders, a local policeman, family members of the victim, and neighbors. Consistently, however, there is a process of legitimation of the use of mobbing. The same picture emerges from the concentrated cases of keroyokan in Karangasem in Bali and Indramayu in West Java. This pattern points to the critical role of social networks for legitimation of violence. Mobbing emerges in areas where local actors approve this phenomenon.

The second notable pattern that emerges from the spatial variation is the concentration of particular forms of mobbing in specific areas. This was noted in the discussion of concentrated cases of local custom in Karangasem (Bali), and gang warfare in Indramayu (West Java). These concentrations suggest that patterns of violence are repeated locally, creating a historical imprint of violence in one locality compared with another. The individual cases described, however, also reinforce this picture. In South Kalimantan, the use of mobbing to respond to adultery--to the violation of local norms--is repeated in the reported cases of keroyokan. According to the dataset, this practice is higher in this province and in these districts of South Kalimantan than anywhere else in the studied provinces. The concentration of witchcraft cases is also echoed in areas south of Bogor, although witchcraft cases are commonly found throughout reported keroyokan in southern and rural West Java, suggesting broader acceptance of mobbing to remove witch doctors. Similarly, burning victims were more common in West Java and Indramayu, where during the period 2001-2002 the most burning victims were reported.

The findings of concentrated forms of keroyokan point to the important role of local learning in shaping concentrations of mobbing. The learning is along two dimensions. First, certain catalysts are accepted and connected with mobbing. Whether it is witchcraft in southern West Java or custom in northeast Bali, the historical accepted link between keroyokan and certain forms of behavior is socialized. Second, the elements in the ritual are also absorbed through learning. In the case of Shaiful in Indramayu, the violent entrepreneur Osman learned the means to kill Shaiful--by setting him on fire--from newspaper accounts and from his repeated participation in massa. This helps explain why burning is more common in this region and why mobbing is more frequent in this area. More opportunities to practice this behavior at the individual level lead to more mobbing, particularly of a specific form.

Taken together, the impact of decentralization, social tacit approval of local leaders, and learned behavior provide an explanation of the variation of keroyokan. National and local factors form a synergy in accounting for this form of horizontal "ordinary" violence.


By drawing attention to the variation over time, space, and form of mobbing, this article has shown that a nuanced understanding of this phenomenon is essential for minimizing this form of local violence in the future. Mobbing has persisted for a long time in Indonesia, and its prominence has increased as horizontal violence has become more common than state-induced conflict. How Indonesia compares with other countries in the practice of mobbing is not known, since the ritual is hard to assess systematically. Yet, research shows that mobbing is a common phenomenon from Africa to Asia. By recognizing the critical role that local and national conditions play in shaping this phenomenon and by moving away from broad generalizations regarding national causes of this phenomenon, a richer understanding of the practice of mobbing can be attained. It is only through combining methods and approaches that the lessons to prevent future violence can emerge.


Abuza, Zachary. 2003. Militant Islam in Southeast Asia: Crucible of Terror. Boulder: Lynne Rienner.

--. 2007. Political Islam and Violence in Indonesia. London: Routledge.

Acciaioli, Greg. 2000. "Ground of Conflict, Idioms of Harmony: Custom, Religion, and Nationalism in Violence Avoidance in Lindu Plain, Central Sulawesi." Indonesia 72:81-114.

Adeney-Risakotta, F. R. 2005. "Politics, Ritual and Identity in Indonesia: A Moluccan History of Religion and Social Conflict." PhD diss., Radbout University.

Anderson, Benedict R. 1989. "Reading 'Revenge' by Pramoedya Ananta Toer (1978-1982)." In Writing on the Tongue, ed. A. L. Becker, 38-83. Ann Arbor: Center for South and Southeast Asian Studies, University of Michigan.

--, ed. 2001. Violence and the State in Suharto's Indonesia. Ithaca: Cornell University, Southeast Asia Program Publications.

Aragon, L. V. 2000. "Elite Competition in Central Sulawesi." In Renegotiating Boundaries: Local Politics in Post-Suharto Indonesia, ed. H. Schulte-Nordholt and G. van Klinken. Leiden: KITLV Press.

Aspinall, Edward, and Greg Fealy, eds. 2003. Local Power and Politics in Indonesia: Decentralisation and Democratisation. Singapore: ISEAS.

Auyero, Javier. 2007. Routine Politics and Violence in Argentina: The Grey Zone of State Power. New York: Cambridge University Press.

Barker, Joshua. 1998. "State of Fear: Controlling the Criminal Contagion in Suharto's New Order." Indonesia 66: 7-42.

--. 1999. "Surveillance and Territoriality in Bandung." In Figures in Criminality in Indonesia, the Philippines and Colonial Vietnam, ed. Vicente L. Rafael, 95-127. Ithaca: Cornell University, Southeast Asia Program Publications.

Barron, Patrick, Claire Smith, and Michael Woolcock. 2004. "Understanding Local Level Conflict in Developing Countries: Theory, Evidence and Implications from Indonesia." Conflict Prevention and Reconstruction Working Paper No. 19. Washington, DC: World Bank.

Baswedan, A. R. 2004. "Political Islam in Indonesia: Present and Future Trajectory." Asian Survey 44: 669-690.

Bertrand, Jacques. 2002. "Legacies of the Authoritarian Past: Religious Violence in Indonesia's Moluccan Islands." Pacific Affairs 75: 57-86.

--. 2004. Nationalism and Ethnic Conflict in Indonesia. Cambridge: Cambridge University Press.

Bjork, Christopher. 2003. "Local Responses to Decentralization Policy in Indonesia.'" Comparative Education Review 47, 2:184-216.

Bourchier, David. 1990. "Crime, Law and State Authority in Indonesia." In State and Civil Society in Indonesia, ed. Arief Budiman, 117-214. Monash Papers on Southeast Asia No. 22. Melbourne: Monash University.

--. 1999. "Skeletons, Vigilantes and the Armed Forces' Fall from Grace." In Reformasi: Crisis and Change in Indonesia, ed. A. Budiman, B. Hatley, and D. Kingbury. Melbourne: Monash Asia Institute, 1999.

Brown, Jason. 2000. "The Banyuwangi Murders." Inside Indonesia (April-June).

Campbell, C., and L. H. Connor. 2000. "Sorcery, Modernity and Social Transformation in Banyuwangi, East Java." Review of Indonesian and Malaysian Affairs 34:61-98.

Colombijn, Freek. 2002. "Maling, Maling! The Lynching of Petty Criminals." In Roots of Violence in Indonesia: Contemporary Violence in Historical Perspective, ed. Freed Colombijn and J. Thomas Lindblad, 299-329. London: KITLV Press.

Colombijn, Freek, and J. Thomas Lindblad, eds. 2002. Roots of Violence in Indonesia: Contemporary Violence in Historical Perspective. London: KITLV Press.

Coppel, Charles A., ed. 2006. Violent Conflicts in Indonesia: Analysis, Representation and Resolution. London: Routledge-Curzon.

Cribb, Robert, ed. 1990. The Indonesian Killings: Studies from Java and Bali. Melbourne: Centre of Southeast Asian Studies, Monash University.

Davidson, Jamie. 2003. "The Politics of Violence on an Indonesian Periphery." South East Asian Research 11, 1: 59-90.

Djalal, Dini. 2000. "The New Face of Indonesian Justice." Far Eastern Economic Review, July 13.

Emmerson, Donald K. 1999. "Voting and Violence: Indonesia and East Timor." In Indonesia Beyond Suharto: Polity, Economy, Society, Transition, ed. Donald Emmerson. Armonk, NY: M. E. Sharpe.

Fealy, Greg. 2002. "is Indonesia a Terrorist Base? The Gulf Between Rhetoric and Evidence Is Wide." Inside Indonesia (July-September).

Foreign Policy. 2006. The Failed States Index 2006. Washington, DC: Foreign Policy.

Goenawan Mohamad. 2002. Conversations with Difference: Essays from Tempo Magazine. Jakarta: Tempo.

Good, Byron J., and Mary-Jo DelVecchio Good. 2001. "Why Do the Masses So Easily Run Amok?" Latitudes 5: 10-19.

Gould, R. V. 1999. "Collective Violence and Group Solidarity: Evidence from a Feuding Society." American Sociological Review 64: 356-380.

Green, Donald P., Jack Glaser, and Andrew Rich. 1998. "From Lynching to Gay Bashing: The Elusive Connections Between Economic Conditions and Hate Crime." Journal of Personality and Social Psychology 75: 82-92.

Gunaratna, R. 2002. Inside Al-Qaeda: Global Network of Terror. New York: Columbia University Press.

Gurr, Ted. 1968. "A Causal Model of Civil Strife: A Comparative Analysis Using New Indices." American Political Science Review 62, 4:1104-1124.

Herriman, Nicholas. 2006. "The Killings of Alleged Sorcerers in South Malang: Conspiracy, Ninjas, or 'Community Justice'?" In Violent Conflicts in Indonesia: Analysis, Representation and Resolution, ed. Charles A. Coppel. London: RoutledgeCurzon.

Hidayat, Syarif. 2007. Too Much Too Soon: Local State Elite's Perspective on and the Puzzle of Contemporary. Indonesia Regional Autonomy Policy. Jakarta: PT RajaGrafindo Persada.

Hinton, Alexander Laban. 1998. "A Head for an Eye: Revenge in Cambodian Genocide." American Ethnologist 25: 352-377.

Human Rights Watch. 1997. Communal Violence in West Kalimantan. Washington, DC: Human Rights Watch, (accessed October 31, 2007).

--. 1999. The Violence in Ambon. Washington, DC: Human Rights Watch Asia, (accessed October 31, 2007).

--. 2001a. Indonesia: The War in Aceh. New York: Human Rights Watch, (accessed October 31, 2007).

--. 200lb. Indonesia: The Violence in Central Kalimantan (Borneo). New York: Human Rights Watch, .htm (accessed October 31, 2007).

--. 2002. Breakdown: Four Years of Communal Violence in Central Sulawesi. New York: Human Rights Watch, indonesia (accessed October 31, 2007).

International Crisis Group (ICG). 2000. Indonesia: Overcoming Murder and Chaos in Maluku. Jakarta: ICG, l&id=1458 (accessed October 31, 2007).

--. 2001a. Communal Violence in Indonesia: Lessons from Kalimantan. Jakarta: ICG, (accessed October 31, 2007).

--. 2001b. Indonesia: National Police Reform. ICG Asia Report No. 13. Jakarta: ICG, (accessed October 31, 2007).

--. 2001c. Indonesia: Violence and Radical Muslims. Asia Briefing No. 10. Jakarta: ICG, (accessed February 8, 2008).

--. 2002a. Al-Qaeda in Southeast Asia: The Case of the "Ngruki Network" in Indonesia. Asia Briefing No. 20. Jakarta: ICG, www.crisisgroup .org/home/index.cfm?id=1765&l=l (accessed February 8, 2008).

--. 2002b. Indonesia Backgrounder: How the Jemaah Islamiyah Terrorist Network Operates. Asia Report No. 43. Jakarta: ICG, www.crisisgroup .org/home/index.cfm?id=1397&1=1 (accessed February 8, 2008).

--. 2002c. Tensions on Flores: Local Symptoms of National Problems Kalimantan. Jakarta: ICG, (accessed October 31, 2007).

--. 2004a. Indonesia Backgrounder: Jihad in Central Sulawesi. Asia Report No 74. Jakarta: ICG, 2500&1=1 (accessed October 31, 2007).

--. 2004b. Indonesia Backgrounder: Why Salafism and Terrorism Mostly Don't Mix. Asia Report No. 83. Jakarta: ICG, index.cfm?id=2967&l=l (accessed February 8, 2008).

--. 2004c. Indonesia: Violence Erupts Again in Ambon. Asia Briefing No. 32. Jakarta: ICG, (accessed Octobcr 31, 2007).

--. 2005. Decentralisation and Conflict in Indonesia: The Mumasa Case. Asia Briefing No. 37. Jakarta: ICG, .cfm?id=34(16&l=l (accessed October 31, 2007).

--. 2006. Terrorism in Indonesia: Noordin's Networks. Asia Report No. 114. Jakarta: ICG, (accessed October 31, 2007).

--. 2007. Indonesia: Jemaah Islamiyah's Current Status. Asia Briefing No. 63. Jakarta: ICG, (accessed October 31, 2007).

Joseph, Lauren, Matthew Mahlaer, and Javier Auyero. 2007. New Perspectives in Political Ethnography. New York: Springer.

Kakir, Sudhir. 1996. The Colors of Violence: Cultural Identities, Religion and Conflict. Chicago: University of Chicago Press.

King, Dwight Y. 2002. "Political Reforms, Decentralisation and Democratic Consolidation," (accessed May 23, 2002).

Kingsbury, Damien. 2005. Violence in Between: Conflict and Security in Archipelagic Southeast Asia. Melbourne: Monash Asia Institute.

Lee, J. P. 1999. "Out of Order: The Politics of Modernity in Indonesia." PhD diss., University of Virginia.

Lindsey, Timothy. 2001. "The Criminal State: Premanisme and the New Indonesia." In Indonesia Today: Challenges of History, ed. Grayson Lloyd and Shannon Smith, 283-297. Singapore: Institute of Southeast Asian Studies.

Madden, David, and Patrick Barron. 2002. "Violence and Conflict Resolution in 'Non-Conflict' Regions: The Case of Lampung, Indonesia." World Bank Report. Washington, DC: World Bank.

Meliala, Adrianus. 2000. "Bagiamana Preman dan Premanisme Hidup di Jakarta." Unpublished manuscript.

O'Rourke, Kevin. 2002. Reformasi: The Struggle for Power in Post-Suharto Indonesia. Sydney: Allen & Unwin.

Peluso, N. L., and E. Harwell. 2001. "Territory, Custom, and the Cultural Politics of Ethnic War in West Kalimantan, Indonesia." In Violent Environments, ed. N. L. Peluso and M. Watts. Ithaca: Cornell University Press.

Platt, G. M. 2004. "Unifying Social Movement Theories." Qualitative Sociology 27:107-116.

Ray, D., and G. Goodpaster. 2003. "Indonesian Cecentralization." In Autonomy and Disintegration in Indonesia, ed. D. Kingsbury and H. Aveling. London: RoutledgeCurzon.

Rhode, David. 2001. "Indonesia Unraveling?" Foreign Affairs 80, 4:110-124.

Robinson, Geoffrey. 1995. The Dark Side of Paradise: Political Violence in Bali. Ithaca: Cornell University Press.

Rohdewohld, Rainer. 2003. "Decentralisation and the Indonesian Bureaucracy: Major Changes, Minor Impact?" In Local Power and Politics in Indonesia, ed. Edward Aspinall and Greg Fealy. Singapore: Institute of Southeast Asian Studies.

Rossa, John. 2006. Pretext for Mass Murder: The September 30th Movement and Suharto's Coup d'Etat in Indonesia. Madison: University of Wisconsin.

Ryter, Loren. 1998. "Pemuda Pancasila: The Last Loyalist Free Men of Soeharto's Order?" Indonesia 66 (October): 45-73.

Saad, I. 2001. "Indonesia's Decentralization Policy: The Budget Allocation and Its Implications for the Business Environment." Jakarta: SMERU Research Institute.

Sangaji, A. 2007. "The Military, Militias and Violence in Poso." In Renegotiating Boundaries: Local Politics in Post-Suharto Indonesia, ed. H. Schulte-Nordbolt and G. van Klinken. Leiden: KITLV Press.

Schiller, A., and B. Garang. 2002. "Religion and Inter-ethnic Violence in Indonesia." Journal of Contemporary Asia 32: 244-254.

Schulte-Nordholt, Henk, and Margreet van Till. 1999. "Colonial Criminals in Java, 1870-1910." In Figures in Criminality in Indonesia, the Philippines and Colonial Vietnam, ed. Vicente L. Rafael, 47-69. Ithaca: Cornell University, Southeast Asia Program Publications.

Shiraishi, Takashi. 1997. "Policing the Phantom Underground." Indonesia 63: 1-46.

Sidel, John. 2006. Riots, Pogroms, Jihad: Religious Violence in Indonesia. Ithaca: Cornell University Press.

--. 2007. Islamist Threat in Southeast Asia: A Reassessment. Policy Studies No. 37. Washington, DC: East-West Center. Siegel, James. T. 1998a. A New Criminal Type in Jakarta: Counter-revolution Today. Durham: Duke University Press.

--. 1998b. "Early Thoughts on the Violence of May 13 and 14, 1998 in Jakarta." Indonesia 66 (October): 75-107.

--. 1999. "A New Criminal Type in Jakarta: The Nationalisation of Death." In Figures in Criminality in Indonesia, the Philippines and Colonial Vietnam, ed. Vicente L. Rafael, 193-209. Ithaca: Cornell University, Southeast Asia Program Publications.

--. 2001. "Suharto, Witches." Indonesia 71: 27-79.

SMERU Research Institute. 1998. "Nationwide Data on the Impact of the Crisis." Newsletter No. 2 (December), 199802spec.htm (accessed October 31, 2007).

Smith, C. J. 2005. "The Roots of Violence and Prospects for Reconciliation: A Case Study of Ethnic Conflict in Central Kalimantan, Indonesia." Washington, DC: World Bank, 67 ByDocName/TheRootsofViolenceandProspectsforReconciliationA CaseStudyofEthnicConflictinCentralKalimantanIndonesia/$FILE/WP23_ Web.pdf (accessed October 31, 2007).

Snow, D. A., and R. D. Benford. 1992. "Master Frames and Cycles of Protest." In Frontiers in Social Movement Theory, ed. A. D. Morris and C. M. Mueller. New Haven: Yale University Press.

Spores, John C. 1988. Running Anzok: An Historical Inquiry. Monographs in International Studies, Southeast Asia Series No. 82. Athens: Ohio University.

Stewart, Frances. 2002. "Root Causes of Violent Conflict in Developing Countries." British Medical Journal 324: 342-345, 324/7333/342.pdf (accesscd October 31, 2007).

Supriatma, A. M. T. (forthcoming). "Communal Violence in Indonesia: A Comparative Perspective." PhD diss., Cornell University.

Tadjoeddin, Mohammad Zulfan. 2000. "Anatomy of Social Violence in the Context of Transition: The Case of Indonesia, 1990-2000." Preliminary Report. New York: UNDP.

--. 2004. "Civil Society Engagement and Communal Violence: Reflections of Various Hypotheses in the Context of Indonesia." Politics Administration and Change 42: 1-18.

Tadjoeddin, M. Z., W. I. Suharyo, and S. Mishra. 2003. "Aspiration to Inequality: Regional Disparity and Centre-Regional Conflicts in indonesia." Jakarta: United National Support Facility for indonesian Recovery (UNSFIR), -2003-1-papers/mohammad%20zulfan%20tadjoeddin.pdf (accessed October 31, 2007).

Tambiah, Stanley J. 1997. "Friends, Neighbors, Enemies, Strangers: Aggressor and Victim in Civilian Ethnic Riots." Social Science and Medicine 45: 1177-1188.

Tilly, Charles. 2003. The Politics of Collective Violence. New York: Cambridge University Press.

Tolney, Stewart E., E. M. Beck, and James L. Massey. 1989. "The Power Threat Hypothesis and Black Lynching: 'Whither' the Evidence?" Social Forces 69: 634-640.

Tornquist, O., ed. 2000. Political Violence: Indonesia and India in Comparative Perspective. Oslo: Center for Development and the Environment (SUM), University of Oslo.

Trijono, L., ed. 2004. The Making of Ethnic and Religious Conflicts in Southeast Asia: Cases and Resolutions. Yogyakarta: Center for Security and Peace Studies (CSPS).

Turner, K. 2003. "Myths and Moral Authority in Maluku: The Cases of Ambon." Asian Ethnicity 4:241-263.

Useem, B. 1998. "Breakdown Theories of Collective Action." Annual Review of Sociology 24:215-238.

van Bruinessen, M. 2002. "Genealogies of Islamic Radicalism in Post-Suharto Indonesia." South East Asia Research 10:117-154.

van der Kroef, Justus M. 1986. "Petrus: Patterns of Prophylactic Murder in Indonesia." Asian Survey 25, 1: 745-799.

van Klinken, Gerry. 2001. "The Maluku Wars of 1999: Bringing Society Back In." Indonesia 71: 1-26.

--. 2007. Communal Violence and Democratization in Indonesia: Small Town Wars. London: Routledge.

Varshney, Ashutosh. 2003. Ethnic Conflict and Civil Life: Hindus and Muslims in India. New Haven: Yale University Press.

Welsh, Bridget. 2001. "Globalization, Weak States and the Death Toll in East Asia." In Violence and Politics: Globalizations Paradox, ed. Kenton Worcester, Sally Bermanzohn, and Mark Ungar, 67-89. New York: Routledge.

--. 2003. "Street Justice: Vigilante Justice in West Java." Paper presented at the Contentious Politics Workshop of Columbia University, February.

Wessel, I., and G. Wimhofer, eds. 2001. Violence in Indonesia. Hamburg: Abera-Verl.

Wilkinson, Steven. 2004. Votes and Violence: Electoral Competition and Ethnic Riots in India. Cambridge: Cambridge University Press. Wilson, Ian Douglas. 2006. "Continuity and Change: The Changing Contours of Organized Violence in Post-New Order Indonesia." Critical Asian Studies 38, 2: 265-297.


(1.) See van Klinken (2007) and Supriatma (forthcoming) for the most comprehensive account of ethnic violence. Excellent individual studies of ethnic violence during this period include Adeney-Risakotta (2005), Davidson (2003), Human Rights Watch (1997, 1999, 2001a, 2001b, 2002), International Crisis Group (2000, 2001a, 2004c), Schiller and Garang (2002), Smith (2005), Trijono (2004), Turner (2003), and van Klinken (2001).

(2.) See Javier Auyero (2007) and for a broader discussion of political ethnography, see Joseph, Mahlaer, and Auyero (2007).

(3.) See Spores (1988) for a richer discussion of this practice.

(4.) See Radar Bogor, June 1, 2000. This report was also buttressed with interviews with morgue officials in Bogor in August 2002.

(5.) Interviews in Tanggerang, June 2002.

(6.) The other two datasets collected were from available police records and morgue data, both of which were more spotty and inconsistent than the newspaper accounts.

(7.) The West Java (including Jakarta) data are derived from Kompas (1997-2004), Pos Kota (1997-2004), Pikaran Rakyat (1997-June 2004), Radar Bogor (2000-2004), and Radar Cirebon (2000-May 2004). The South Kalimantan data are gathered from the Banjarmasin Post from 1995 through July 2003. The Bali data are derived from three newspapers: Bali Post (1995-2004), Nusa Post (1995-2004), and Denpasar Post (2000-2003). Only two of these are "tabloids"--Pos Kota and Nusa Post.

(8.) The accuracy of these papers is at times questionable, yet they base their reports on actual facts and sensationalize the reports rather than construct fabricated reports. This conclusion was reached after randomly testing a series of the reports as part of our in-depth interviews of cases in different provinces.

(9.) Pikaran Rakyat, May 3, 1999.

(10.) For an excellent discussion of revenge in Indonesia, see Anderson (1989), and for the similar practice in Cambodia, see Hinton (1998).

(11.) The most vibrant discussion of witchcraft involved the Banyuwangi murders in East Java. See Brown (2000), Campbell and Connor (2000), Herriman (2006), and Siegel (2001).

(12.) The interviews for this research were conducted in October 2004.

(13.) The interviews for this research were conducted in June 2002.

(14.) The interviews for this case study were conducted in July 2003 in the village of Gambut and included the victim Hip, family members of Jako-notably his sister Ana and nephew Mat--Hip's relative Atma, the local priest, and neighbors.

(15.) The interviews for this case study were conducted in August 2002 in the village of Cimanggu Dua with Hussin, his wife Sharifah, neighbors, the local imam, and neighbors.

(16.) The interviews for this case study were conducted in June 2003 in Indramayu and included Mariah, Osman, neighbors in the community, and the local policeman, Hari.

(17.) Two collections cover the range of the explanations below: Colombijn and Lindblad (2002) and Coppel (2006). For an alternative theoretical discussion, see Barron, Smith, and Woolcock (2004).

(18.) See Anderson (2001), Barker (1998), Kroef (1986), Lee (1999), Lindsey (2001), Rossa (2006), Shiraishi (1997), Tornquist (2000), and Wessel and Wimhofer (2001). Most of the discussion in Kinsgbury (2005) also highlights state action.

(19.) See Bourchier (1990, 1999), Djalal (2000), Foreign Policy (2006), International Crisis Group (2001b), Kingsbury (2005), and Rhode (2001).

(20.) For discussion of economic issues, see Green, Glaser, and Rich (1998), SMERU (1998), Steward (2002), and Tadjoeddin, Shaharo, and S. Mishra (2003).

(21.) For two interesting discussions of the relationship between decentralization and conflict, see International Crisis Group (2001c, 2005).

(22.) The most prominent writer using this explanation on Indonesia is van Klinken (2007). See also Tadjoeddin (2004), Platt (2004), Tilly (2003), and Useem (1998).

(23.) See Gould (1999) and Snow and Benford (1992). For the historical discussions of concentrated violence, see Cribb (1990) and Robinson (1995).

(24.) For trauma and threat affecting violence, see Kakir (1996) and Tolney, Beck, and Massey (1989). For a discussion of criminality, see Barker (1999), Meliala (2000), Ryter (1998), Schulte-Nordholt and van Till (1999), and Siegel (1998a, 1999).

(25.) For a psychological discussion of "amok," see Good and Good (2001). For a discussion of the critical role of local custom, see Acciaioli (2000) and Peluso and Harwell (2001).

(26.) For a discussion of local interpersonal dynamics, see Tambiah (1997).

(27.) For a good discussion of decentralization, see Aspinall and Fealy (2003), Bjork (2003), Hidayat (2007), King (2002), Ray and Goodpaster (2003), Rohdewohld (2003), and Saad (2001).

(28.) This discussion is based on interviews with local officials in Bali in October 2004.

Bridget Welsh is assistant professor in the Southeast Asia Studies Program at Johns Hopkins University-SAIS in Washington, DC, where she teaches courses in contemporary Southeast Asian politics. In 2004, she was a Henry R. Luce Southeast Asian Fellow at Australian National University. Her most recent publications include Reflections: The Mahathir Years (2004), Legacy of Engagement in Southeast Asia (2008), and Impressions of the Goh Chok Tong Years in Singapore (2009).
Table 1 Overview of Types of Keroyokan

Catalyst Premeditation Brutality Target

 Petty theft Low Medium/high Male out/insider
 Robbery Medium Medium/high Male out/insider
 Hit-and-run Low Medium/high Male outsider
 Rape Medium Medium/high Male out/insider
Personal acrimony
 Revenge High/medium Medium Male insider
 Gang warfare High/medium Medium Male insider
 Ninja killing Medium Very high Male outsider
Norm violation
 Deviant removal High High Male insider
 Witchcraft High Very high Male/female
Local custom High High Male insider

Catalyst Social Legitimation

 Petty theft Community justice
 Robbery Community justice
 Hit-and-run Community justice
 Rape Community Justice
Personal acrimony
 Revenge Assault/murder
 Gang warfare Assault/murder
 Ninja killing Assault/murder
Norm violation
 Deviant removal Assault/murder
 Witchcraft Assault/murder

Local custom Community justice

Table 2 Keroyokan in Selected Provinces, 1995-2004:
West Java, Bali, Bengkulu, and South Kalimantan

Year Incidents Victims

1995 214 282
1996 126 168
1997 204 318
1998 191 416
1999 270 439
2000 564 795
2001 781 1,007
2002 539 688
2003 566 692
2004 582 701
Total 4,037 5,506

Table 3 Keroyokan by Specific Selected Provinces

Province Incidents Victims

West Java 2118 2,797
Bali 1,206 1,782
Bengkulu 314 391
South Kalimantan 399 536

Table 4 Keroyokan Catalysts

 Form Percentage

Petty theft 47
Extortion 21
Gang warfare 9
Witchcraft 6
Revenge 5
Social deviant 3
Local custom 2
Rape 3
Adultery 2
Hit and run 1
Ninja attacks 1
COPYRIGHT 2008 Lynne Rienner Publishers
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Welsh, Bridget
Publication:Journal of East Asian Studies
Geographic Code:9INDO
Date:Sep 1, 2008
Previous Article:Explaining ethnic violence in Indonesia: demilitarizing domestic security.
Next Article:The East Asian High-Tech Drive.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters