Operational performance of regional electricity distribution in Indonesia

AFRIZAL, Prof. Kaneko Shinji

2008 | Tesis | S2 Magister Ekonomika Pembangunan

After three decades of sustained growth, the Indonesian electricity industry has been enduring a very difficult phase, as an adverse impact of the country’s economic crisis. The effort to improve efficiency has been carried out since 1999. During the transition process of the reform, efforts to promote efficiency in electricity sector were undertaken. Since 2000, the state owned electricity enterprise, Perusahaan Listrik Negara (PLN) has implemented an Efficiency Drive Program (EDP) in the electricity sector. The target of EDP in distribution is to minimize losses and to increase income. This is done by optimizing the operation pattern and maintenance, and by conducting other efficiency improvement activities. Income improvement will also come from regulated tariff adjustments and reduced billing cycle. The EDP is part of the program to meet the government’s vision to provide electricity for all by 2020. To get the overall picture and the effect of EDP, this study evaluates the changes in operational performance of regional electricity distribution in Indonesia. The target companies are 22 regional electricity distributions and the period of observation is from 2002 to 2005. The study measures performance using Data Envelopment Analysis (DEA). DEA is a non parametric model based on mathematical linear programming technique for measuring the relative performance of organizational units using multiple input and output data. Since DEA was first introduced by Charnes et al (1978), this methodology has been widely applied to the measurement of efficiency and productivity. In this study, performance measurements are based on efficiency and Malmquist total factor productivity (TFP). Variables used to calculate performance (efficiency and TFP) are three inp uts and three outputs. The input variables are total number of employees in distribution (person), the length of existing electricity cables (Grid size, km) and transformer capacity (MVA). The output variables are the amount of electricity distributed to end users (unit sold, GWh), total number of customer (connection) and total real revenue or turnover in a year (Rev, Million Rp. Real Price). The relative efficiencies are calculated in static and dynamic DEA under constant return to scale (CRS) assumption. The VRS model is calculated to see the effect of scale on efficiency. The study applies two stages of DEA. In the first stage, the study calculates performance based on DEA efficiency and Malmquist TFP. Efficiency scores are calculated using static and dynamic DEA (output-oriented Model) to get efficiency score and potential improvement. The next step, the Malmquist TFP index is used to calculate the changes in TFP. The calculation is based on dynamic DEA CRS and its decomposition into catch up (technical efficiency change) effect and frontier shift (technological or innovation) effect. In the second stage, the study identifies the efficiency determinants and the Tobit regression is applied in this stage. The impact of losses, price and other determinants are considered. Tobit model is employed when the dependent variable is censored and limited (here efficiency score from 0 to 1) and the scores are from dynamic DEA (Window Analysis). The results show that there was performance improvement from 2002 to 2005 based on efficiency and TFP measurement. Efficiency measurements increase during the study period using static and dynamic DEA. The efficiency scores indicate the performance level of company. In aggregate mean efficiency, the scores from dynamic DEA are slightly less than in static DEA. The dynamic measurement can give better indication for improvement since the score is lower than static DEA. The distribution sector has an average aggregate efficiency of 83.8 percent in over 4 years from dynamic DEA. Performance improvement is also shown by Malmquist TFP measurement. The change in TFP over 4 years is 13.4 percent. This value is induced by 5.9 percent of efficiency change and 7 percent of technical change. On average, there is 4.4 percent growth rate in TFP over 4 years. The value is good since the sector has a limitation in investment ability. The performance measurement shows the positive effect of EDP. The results of efficiency analysis also indicate by how much and in what areas an inefficient unit needs to improve in order to achieve maximum efficiency. There are possibilities to improve the efficiency by increasing units sold in 6.07 percent, customer in 8.05 percent and revenues in 6.15 percent using dynamic DEA. From the government or regulator side, inefficiencies of customer indicate the possibility to increase electrification ratio. The econometric exercise from Tobit model provides evidence regarding the determinants of efficiency scores. The results show that losses index, GDP per capita, electrification ratio, average price and Java dummy appear to significantly affect the efficiency. Meanwhile, the average price did not show the expected sign. As a target of efficiency drive program, the losses index has adverse effect on performance. The result is consistent with the objective of EDP. It is suggested that EDP program decreases losses if improved and regularly evaluated. Positive economic growth shown by the improvement of GDP per capita suggests that the government should improve the economy in order to improve the performance of the sector, which is already in line with the government’s vision concerning the electricity sector in Indonesia. Java Dummy which captures business specialization indicates that restructuring the corporate might be needed to separate the system of generation, transmission and distribution outside Java. The opposite sign of average price might be influenced by the uniform tariff application, and the company considered as benchmark did not show improvements in their ave rage price. This uniform tariff across the country should be reviewed. There should be tariff differentiation based on population density, location and economy (regional income per capita). Government should design a new legal framework to support the restructuring and regional tariff policy. The study also recommends a challenge to conduct further study which incorporates cost data to measure allocative or cost efficiency. The econometric exercise from Tobit model provides evidence regarding the determinants of efficiency scores. The results show that losses index, GDP per capita, electrification ratio, average price and Java dummy appear to significantly affect the efficiency. Meanwhile, the average price did not show the expected sign. As a target of efficiency drive program, the losses index has adverse effect on performance. The result is consistent with the objective of EDP. It is suggested that EDP program decreases losses if improved and regularly evaluated. Positive economic growth shown by the improvement of GDP per capita suggests that the government should improve the economy in order to improve the performance of the sector, which is already in line with the government’s vision concerning the electricity sector in Indonesia. Java Dummy which captures business specialization indicates that restructuring the corporate might be needed to separate the system of generation, transmission and distribution outside Java. The opposite sign of average price might be influenced by the uniform tariff application, and the company considered as benchmark did not show improvements in their ave rage price. This uniform tariff across the country should be reviewed. There should be tariff differentiation based on population density, location and economy (regional income per capita). Government should design a new legal framework to support the restructuring and regional tariff policy. The study also recommends a challenge to conduct further study which incorporates cost data to measure allocative or cost efficiency.

Kata Kunci : Distribusi,Elektrik industri performan,Regional efficiency