Data & Papers › Dataset
Indonesian Ethnic and Religious Coalitions
With Edward Aspinall, I collected an extensive record of ethnic and religious coalitions in Indonesian regional head elections: 675 candidate tickets in 123 elections across seven provinces, 2010–2013. We combined it with a survey of over 500 politicians and interviews with campaign workers to explain why candidates from different groups join forces, and why many build broader coalitions than they need to win.
About the data
- Elections
- 123 regional head elections (governors, mayors and regents) across seven provinces, 2010–2013.
- Tickets
- 675 candidate tickets (a head and a deputy candidate), with the ethnic and religious identity of the candidates.
- Fieldwork
- A survey of over 500 politicians and interviews with campaign workers.
- With
- Edward Aspinall (Australian National University).
- Materials
- Variable descriptions are in the replication package; this dataset has no separate codebook, as it is not a content analysis.
Papers using the data
- Forth.
Beyond Demographics: Ethnic and Religious Coalition Formation in Indonesia
With Edward Aspinall. Perspectives on Politics.
Existing research on intergroup coalition formation in diverse democracies emphasizes demographic calculations. The most influential approach predicts that candidates will form “minimum winning coalitions” just large enough to secure victory while minimizing their patronage obligations. To address this topic, we collected a unique dataset on interreligious and interethnic coalition formation by almost 700 Indonesian candidates competing in 123 local elections, surveyed over 500 politicians, and interviewed campaign workers. We found that intergroup coalition formation is explained by two distinct strategic logics. The first is size-based: candidates from smaller groups form intergroup coalitions to assemble a winning vote share. The second is status-based: candidates from groups that are vulnerable in national or regional hierarchies form coalitions with dominant groups in order to signal legitimacy and shield against backlash. Crucially, this status logic operates even when candidates from vulnerable groups already have local winning majorities. Finally, we found that candidates tend to maximize rather than minimize their intergroup coalition size, with scant evidence that candidates form minimum winning coalitions.
Using and citing the data
The data and code are published as a replication package on Harvard Dataverse. Please cite the article and the replication package; the Dataverse page gives the full data citation.
Bahasa Indonesia
Bersama Edward Aspinall, saya menghimpun catatan koalisi etnis dan agama dalam pilkada: 675 pasangan calon dalam 123 pemilihan di tujuh provinsi, 2010–2013, dilengkapi survei terhadap lebih dari 500 politisi dan wawancara dengan tim kampanye.