DEKOMPOSISI PERTUMBUHAN PRODUKTIVITAS INDUSTRI MANUFAKTUR INDONESIA: PENDEKATAN STOCHASTIC FRONTIER

Abstract

Productivity growth can be decomposed into its sources, namely technical changes, technical efficiency changes, and business scale changes. The role of each of these sources is important to be studied because Indonesia experienced a decline in total factor productivity (TFP) growth in 1990-2015 which indicates a productivity issue. The results of previous studies by Margono Sharma (2006) and Ikhsan (2007) regarding the decomposition of productivity growth in the manufacturing industry sector in Indonesia using a stochastic frontier approach, there are still disagreement about the contribution of technical changes and technical efficiency changes to TFP growth. This is thought to be caused by differences in the specifications of the technical efficiency models used, namely Battese Coelli (1992) and Lee Schmidt (1993). The two models have different estimation methods, Battese Coelli (1992) use maximum likelihood estimate, while Lee Schmidt (1993) use non-linear least squares. Hence the scalar parameters that determine the pattern of technical efficiency over time (η in Battese Coelli (1992) and θ in Lee Schmidt (1993)) are different.This study uses data for the period of 2003-2014 in the manufacturing industry sector. The results of this study indicate that different technical efficiency model specifications produce a different relative contribution of TFP growth sources. Therefore, choosing a technical efficiency model to decompose TFP growth should be thorough if the results will be used in important policy making.