Since 2001, Indonesia has undergone wide-ranging decentralization. Political power and administrative responsibility have been transferred from the central government to the district level. District authorities now have the political mandate to issue decrees and regulations and full responsibility for the public management of sectors such as education, health care, public works, culture and the environment. The devolution process follows the economic crisis which began in 1997, the fall of the authoritarian regime of president Suharto in 1998, and the implementation of free and democratic elections in 1999.
There is a long tradition of people and groups taking the law into their own hands in Indonesia. However, there has been an increase in horizontal violence after the change of political leadership and structures in 1998/99. Theft and destruction of private property has become common and large numbers of ordinary people in many parts of the country have been threatened or suffered from physical violence or been tortured and even executed by their peers. Nationwide, several hundred people have been killed every year by angry mobs in incidents of street vigilantism (Kompos, 25 September 2001). Much of the horizontal violence observed over the last few years is related to the economic crisis and unemployment, combined with the weakening of central state institutions, including the police. Lasting problems of horizontal violence are also often associated with ethnicity, as in Solo, Mataram and Medan, and religion, as in the Moluccas and Central Sulawesi. The police force has not been decentralized. After being separated from the military in 1999, the national police has been administratively placed directly under the authority of the president, with some responsibilities still delegated to the provincial governors. The decentralization laws and regulations do not delineate specific roles for the police and security matters at the district level. This means that the delegation of huge responsibilities to the districts is not formally backed up by law enforcement power. Instead, we see an alarming development of local security forces and paramilitary groups controlled by district authorities. These security forces can hardly be interpreted as anything but an instrument to create legitimacy by force, instead of by popular support or democratic procedure, for local government policies.
The main objective of this article is to trace the impact of decentralization reforms on law enforcement and security systems at the district level. Four administrative regions which have different experiences in security problems and organizing paramilitary groups were selected for study. Two questions are raised: have decentralization reforms strengthened the tendencies to escalating disorder and crime rates in Indonesian society? What are the impacts of formally separating law-making and -enforcing systems in the decentralization reforms? The methodology used is mostly qualitative, based on in-depth interviews and focus group discussions. In addition, we raised questions regarding security problems and systems in a survey that included more than 500 households in the four districts.
This article is organized as follows: the introduction is followed by an overview of the history of criminality, law enforcement and security systems in Indonesia. The introduction of the new decentralization laws and their implementation since 2001 are then discussed. Methodology and research areas are presented in section four and our empirical findings are discussed in section five. A conclusion and policy recommendations are given at the end of the article.
Informally organized security systems have long existed in Indonesia. It was not until 1900 that reasonably standardized police forces appeared within the Dutch colony. Till then, most of the non-European quarters of the colony's cities and towns were "policed" by volunteer neighbourhood watches, known as rondo, who routinely treated suspected thieves, burglars, and other undesirables with vigilante violence (Anderson 2001, p. …