Regional productivity growth in Indonesia: a DEA Malmquist productivity index analysis

Accurate assessment of regional productivity growth is important for policy making. However, most of the productivity growth analyses of Indonesia are conducted at the national level. This thesis presents a comprehensive study of Indonesian regional productivity growth. It covers twenty-six provinces during the period 1985-2010. It aims to make several contributions. First, due to the unavailability of regional capital stock data, this research will construct a new database of Indonesian regional capital stock. Second, most of the previous studies of Indonesia’s productivity growth followed the growth accounting method. This study for the first time decomposes productivity growth at the regional level in Indonesia by using the DEA Malmquist productivity index (MPI) approach. Third, the examination applies different versions of the DEA-MPI (namely the conventional, fixed base year, sequential and metafrontier) and compares the results. Fourth, it provides probably the first study comparing regional productivity growth between Indonesia and China which will enrich the cross-country comparison literature. Finally, the productivity growth decomposition in this study makes it possible to examine the sources of productivity convergence or divergence in Indonesia’s regions.
The estimation results show that most of the Indonesian regions experienced productivity growth. On average, regional productivity growth was more than one per cent. The dominant factor of regional productivity growth was efficiency change which was offset by technical regress. However, estimation based on the islands shows that Java-Bali and Sulawesi islands experienced productivity growth which was dominated by technical change. In addition, based on technological gap estimates, Indonesian regions can be classified into three groups. These are regions that were very close to the global (national) frontier but experienced an inward shift of their group frontier, regions that were relatively close to the global (national) frontier and experienced an outward shift of their group frontier and regions that were far from the global frontier but experienced an outward shift of their group frontier. Regional productivity growth comparison shows that regional productivity growth in Indonesia was higher than that in China and it was dominated by efficiency change whilst China’s was dominated by technical change. Three convergence tests, namely sigma, absolute beta and conditional beta convergence tests, support the existence of productivity convergence in Indonesian regions. The tests also reveal that efficiency change is the main source of regional productivity convergence while technical change is the source of divergence.
Overall, this study of productivity growth at the regional level in Indonesia sheds more light on regional development. It fills the gap in the literature on regional productivity growth in Indonesia by employing the DEA-MPI technique and decomposing productivity growth into several components. This topic becomes important for the discussion of the decentralization policy. Therefore, the findings may have policy implications for both national and regional development in Indonesia.