Poster Paper: When Extractive Political Institutions Affect Public-Private Contracting: Empirical Evidence from Indonesia’s Independent Power Producers Under Two Political Regimes

Monday, July 29, 2019
Indoor Courtyard - Level -1 (Universitat Pompeu Fabra)

*Names in bold indicate Presenter

Yohanna Gultom, Universitas Indonesia


This paper examines the relationship between political institutions and the economic performance of public-private contracting in the provision of public goods. This paper test how extractive political institutions are associated with the performance of public-private contracting using the case of the electricity sector in Indonesia, where a political regime shift has impacted public-private contracting arrangement in the power generation sector. Using two-stage empirical research, namely data envelopment analysis and the difference-in-differences regression, I examine the economic performance of 20 coal-fired plants for the period of 2010-2016 that consists of the independent power producers (IPPs) endorsed by the two political regimes, the authoritarian and the democratic governments, and the state-owned power plants. The results indicate the extractive political institutions are associated with reduced efficiency of the first generation of IPPs which endorsed by the Soeharto regime by –0.135 points, or 0.16% of the mean. This finding shows that across all other power plants that produce the same outputs, the plants that endorsed by the Soeharto regime use inputs or expenditures 0.16% higher than the average of inputs or expenditures used by other plants. The findings challenge the assumption that a transaction cost economizing motive might underlie the use of public-private contracts in power generation sector in Indonesia under the authoritarian regime. Instead, these findings are consistent with the political economy argument that extractive political institutions might have created economic policies that allow for the political elite to extract rents from public-private contracting.