Entitlement programmes for garnering votes? The impact of local elections on targeted social protection in decentralised Indonesia

Purpose - The purpose of this paper is to find the relationship between local direct elections and the change in social spending, controlling for GDP per capita, revenues and wide-range socio-welfare indicators at the regional level. Design/methodology/approach - This research uses a model of time-series cross-sectional panel data set for 33 provinces in Indonesia from 2001 to 2012. Findings - The main finding of this research is that the political budget cycle does exist in Indonesia. Incumbents responded to the direct elections more sensitively rather than to other variables in the model. The most important variables that are significant in the model are not only direct election but also inter-governmental grants. Interestingly, the local economy (as measured by GDP per capita) does not clearly exhibit a meaningful impact. Research limitations/implications - Although the importance of decentralisation in Indonesia is actually at the regency level, obtaining the data is really challenging. Therefore the exercise on this paper is currently limited only for the provincial level. Practical implications - This finding conveys the message that there is large room for improvement in inter-governmental transfer formulation, more importantly to the regions where they still entail significant budget support from central government. In addition, transfers during specific periods such as elections need to be modified to avoid the misappropriation of local budget and to mitigate the adverse impact of PBC. The formulation of inter-governmental transfers is pivotal in reducing over-dependence to the central government funding and to ensure the effectiveness of budget devolved at the local level. Originality/value - To the author’s understanding, the paper is the first to discuss the presence of the political budget cycle on social protection programs in Indonesia. The expected contribution of the current work is twofold: Firstly, the author used a recent data set hosted by the World Bank (INDO-DAPOER). Secondly, the findings are relevant to the discussion within the sphere of development studies and political science.