Essays on decentralisation and accountability in Indonesia

Abstract

The first paper examines the degree of spatial interdependence in fiscal policy and service provision by local governments in Indonesia. The research forms the first attempt to test such interdependence across local areas of a developing country. A spatial Durbin model is estimated for a balanced panel of Indonesian districts for 2009-2013. The findings confirm the existence of positive spatial interdependence in both local government expenditure and access to services. Interdependence is stronger for capital-based spending and education access. The findings point to strategic competition between districts, a process that potentially boosts local government accountability and efficiency in providing service access for citizens.

The second paper contrasts two distinctive government forms at the village level in Indonesia: desa, which is a more citizen-based government with directly elected village heads and kelurahan, which is a bureaucratic-based government with appointed village heads. The paper investigates if switching to the more bureaucratic type of governance influences service access outcomes. Both village-level panel Difference in Difference (DID) for village level outcomes and pseudo-village panel DID for individual outcomes are pursued. The results show that kelurahan status leads to an increase in property crime incidence and reduced access to safe birth and immunisation. These findings point to the importance of horizontal accountability and village apparatus engagement, while also providing insights into the debate on whether political decentralisation should be extended to the lowest administrative tier.

The last paper investigates the consequences of electing a female leader on fiscal and service outcomes. A dataset of districts in Indonesia with close election results between male and female mayoral candidates from 2005 to 2017 is used. This study employs a randomisation-based inference in regression discontinuity design to deal with non-random assignment of female leadership. The results show that per capita expenditure, as well as budget share on social protection and infrastructure, are higher for districts that are governed by female leaders. Also, female leadership improves citizen access to both assisted birth by health professionals and safe water. Finally, the findings show that female-led districts have more prudent budget management while neither fewer nor more corruption cases. This study suggests that female leader policy preferences are different from those of their male counterparts, and that gender plays a role in determining policy choices.