The Causal Effect of Local Elections on Economic Growth- Using a Natural Experiment in Indonesia

Does democracy increase economic growth? This thesis has narrowed down the question to empirically estimating the causal effect of local elections of district heads on economic growth in Indonesia by using a quasi-experimental research method, or “natural experi-ment ” approach. In Indonesia the first elections of district heads were performed in a staggered manner, and decided such that the year of election is as good as exogenous, or random. Thus, the districts that held their first elections in 2005 can be compared with the districts that held their first elections in 2008 or later, which more specifically is performed by using difference-in-difference estimation. The research methodology is innovative and more reliable than previous attempts to capture the causal effect of democracy on growth. The empirical assessment indicates an insignificant effect of local elections on economic growth, except for a significant and positive effect of local elections on growth for the districts of the two more economically- advanced islands, Java and Bali. The main reason behind the insignificant result presented is thought to be heterogeneous effects, that districts react differently to local elections dependent on initial characteristics. However, a limitation is that the time horizon investigated is the effect on growth rates three to five years after elections. Hopefully the results can serve as a reference case for other developing countries, and guide ongoing reform initiatives in Indonesia