Vigilantism and the transition to local democratic elections

ABSTRACT

I examine the impact of the transition to local democratic elections on vigilantism in Indonesia. Using an event study model, I find that the introduction of local elections led to an increase in vigilante conflict, consistent with most theory and qualitative evidence focussing on national level impacts. I determine that the effects are particularly noteworthy in long-established districts and that the impacts are positively associated with rising state-led violence in law enforcement, results that argue against the propositions that lack of state capacity leads to rising vigilantism and that vigilantism serves to substitute for the state’s weak enforcement of law. Contrary to other research, I find no evidence of increased vigilantism in the run-up to direct local elections, and I establish that vigilante conflict appears to be short-lived after the initiation of local elections, a determination that helps to resolve another debate in the literature.

KEYWORDS:

• Democratic transition
• vigilantism
• local government
• Indonesia

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/03003930.2022.2103673

Disclosure Statement

No potential conflict of interest was reported by the authors.

Notes

1 The definition is from Jaffrey (2021). For similar definitions see Moncada (2017) and Bateson (2021). The various definitions are discussed in some detail in the subsequent section of the paper.

2 Pierskalla and Sacks (2017) investigate the association of direct local elections with vigilante and other types of routine conflict in Indonesia but do not identify causal effects of elections on vigilantism. Tajima (2013) examines the causal effects of prior exposure to military interventions on general violence, but not vigilantism, at the village level during Indonesia’s transition to democracy.

3 I have reordered these concepts from how they appear in Moncada (2017) and provided my own interpretations in some cases.

4 For a thorough account of Indonesia’s patronage democracy see Aspinall and Berenschot (2019).

5 For case studies of mass organization (ormas) vigilante violence in Indonesia see Welsh (2008) and Tyson (2013).

6 In 1999 the MPR comprised 500 DPR legislators and 200 additional appointed members drawn from the military and selected civilians.

7 There are also 78 “singleton” districts in the full sample, and these are deleted from estimation, as usual, leaving 3,287 observations in the total estimating sample.

8 An alternative outcome variable would be the annual number of violent events associated with vigilantism. But the size distribution of this variable is very skewed to the right, creating outlier challenges for the regression analyses. Nevertheless, I provide a robustness test to the main results using the number of vigilante events as the outcome of interest and this is discussed later in the paper.

9 The absence of pre-trends does not necessarily guarantee parallel trends in the post-treatment period, in absence of treatment, but it would be very difficult to argue in favor of post-treatment parallel trends if trends were not parallel in pre-treatment period.

10 I define the “last treated” group to be those districts in the sample at years greater than or equal to 2012.

11 See Callaway and Sant’Anna (2021) for an example of estimation under the assumption that anticipatory effects are present.

12 It would be useful to estimate the impact of direct elections on mother and child district separately, especially since the transition to directly elected mayors is qualitatively different across these two cases, (i.e., from council-appointed mayors in the first instance and from Ministry of Home Affairs appointed heads in the latter case) but there are an insufficient number of observations to accommodate this approach.

13 These data also come from SNPK. The latter defines law enforcement violence as “violent action taken by members of formal security forces to perform law-enforcement functions, including use of violence mandated by law as well as violence that exceeds mandate, for example, torture or extra-judicial shooting”.

14 I do not examine the impact of direct elections on separatist violence here as this is the subject of other, ongoing work. Neither do I consider the potential effects of elections on crime and domestic violence as these types of conflict are quite different in kind to those considered here.

Additional information

Notes on contributors

Blane D. Lewis

Blane D. Lewis is Professor of Economics in the Arndt-Corden Department of Economics, Crawford School of Public Policy, Australian National University. He is Director of ANU’s Indonesia Project and Lead Editor of the Bulletin of Indonesian Economic Studies. His research focuses on intergovernmental fiscal relations, local public finance and public service delivery, and local political economy in Indonesia.