Market Reforms and The Performance of The Malaysian Electricity Industry

Electricity market reforms have occurred in many countries, including both developed and developing countries. The main objectives for electricity market reform in developed countries focus on reducing the production costs and electricity prices while the developing countries aim to improve quality of services, mobilise finance and expand electrification (Woolf & Halpern, 2001). Different countries have adopted a variety of market structures for their electricity industries. In general, the choice of which market liberalisation structure to adopt, e.g. monopoly, single buyer, wholesale competition or retail competition, is in part influenced by the state of their economies and their development levels. However, there is no guarantee that market reforms will ultimately bring the projected benefits to consumers and economy. This is because there is no single market structure that is ideal in all situations (Joskow, 2006; Sioshansi, 2008). In the case of Malaysia, the electricity industry has been undergoing regulatory reforms since the mid 1990’s. The generation sector has been partially liberalised by allowing private entry into the market. However, almost all power is purchased by the government-owned company Tenaga Nasional Berhad (TNB) under long term power purchase agreements (PPAs). In the late 1990’s, there was some discussion of the possibility of creating a power pool (similar to those seen in a number of developed countries) to increase competition, rather than endorsing these long term purchase contracts. If the Malaysian government wishes to adopt a more competitive market structure, an evaluation of the performance of electricity industry in terms of the technical efficiency and total factor productivity growth could be useful and valuable in drafting any future policy. Since the early 1990s, many countries have undergone market reform in their electricity industry. It is not surprising that the impacts of market reforms and various environmental variables (i.e., ownership, plant age, etc.) have been investigated in a number of efficiency and productivity analysis studies. The majority of empirical studies have involved applications to United States data. Yunos & Hawdon (1997) is the only empirical study (as far as we are aware) that measured the technical efficiency of the Malaysian electricity industry. Unfortunately, their data is rather dated, and hence does not reflect the current situation in Malaysia. Also, there is no published research (that we know of) that specifically evaluates the technical efficiency (TE) and total factor productivity (TFP) growth of electricity utilities in Malaysia using a parametric model. Hence, in this study we assemble three new data sets and conduct a comprehensive analysis of the TE and TFP growth of electricity utilities in Malaysia. This thesis provides the first comprehensive study (to our knowledge) of the performance of the Malaysian electricity industry, which investigates the relative technical efficiency and TFP growth of electricity utilities. There are three separate empirical analyses conducted; and the findings have important policy implications and provide valuable empirical information that can be used to help guide deliberations regarding future reform policies. We now briefly summarise each of these studies in turn. The first empirical study examines the TFP growth of TNB (and the Malaysian electricity industry as a whole) over a 30 year period from 1975 to 2005 using the Törnqvist index method. The model involves four inputs (i.e., undepreciated replacement value of capital stock, full time employees, fuel consumption and other inputs) and three outputs (i.e., electricity generated, residential and non residential electricity delivered). The results indicate that TNB obtains an average annual TFP change of 0.53 percent while the industry (TNB plus IPPs) achieves an average TFP change of 0.13 percent per year. The introduction of private sector involvement in the Malaysian electricity industry in the 1990’s is observed to coincide with lower TFP growth rates. Furthermore, a sensitivity analysis is conducted where two different capital stock measures (the perpetual inventory method and undepreciated real capital stock) are used, and are found to have a limited influence upon on the empirical results obtained. Our results could suggest that the partial privatisation of TNB and introduction of private entry were insufficient to produce improved TFP performance. Based on experiences in other countries, we hypothesise that the introduction of a competitive market and/or a regulatory framework that encourages more efficient behaviour may also be required. Our empirical findings will hopefully provide useful input into any future discussions of new policy directions in the electricity industry in Malaysia. The second empirical study focuses on the efficiency and productivity differences between TNB-owned and privately-owned power plants in Malaysia. The stochastic frontier and technical inefficiency effects model (Battese & Coelli, 1995) was applied to measure TE levels and TFP growth rates in the Malaysian electricity generation sector. The data involved 86 observations for 14 Malaysian thermal power plants that were collected from 1998 to 2005. The four input variables used in the model were undepreciated replacement value of capital stock, labour, fuel, and other inputs, with electricity sent-out as the output measure. The empirical results show that there still a room for the TNB-owned power plants and IPPs to improve their performance (i.e., average TE score is 79 percent). The result suggests that the average plant would have to increase their technical efficiency level by 21 percent in order to reach the efficient production frontier. Furthermore, the results also show that power plants under public ownership are not performing as well as the privately-owned power plants in Malaysia. Two possible explanations for this stand out. Firstly, if the capacity of TNB owned power plants is not as fully utilised as the IPPs in Malaysia (because of the generous nature of the Power Purchase Agreements), an increase in capacity utilisation may be sufficient to improve TNB’s plant efficiency. Alternatively, the IPPs may be more efficient because the quality of management is higher. If this is the case, privatising TNB’s power plants (in association with a more competitive market) may be considered. In addition to investigating the effects of ownership on technical efficiency levels, our SFA analysis also finds that plant size and fuel type have a significant influence upon the technical efficiency of power plants in Malaysia. This information may be useful in future planning relating to the promotion of improved plant efficiency in the industry. The SFA model is also used to measure productivity growth for the electricity generation industry over the 1998 to 2005 sample period. We find that it achieved average annual TFP growth of 2.34 percent, with technical change contributing the most to the TFP growth over the eight year period. We hence hypothesise that the new power plants with their newer capital-embodied technologies commencing during the sample period are likely to be the main reason for this strong technical change. The third empirical study is used to identify the performance difference between Malaysian and international electricity generating companies. The data involved 190 annual observations across 65 electricity generating companies in six different Asia Pacific countries (the United States, Australia, Singapore, South Korea, Thailand and Malaysia) that were collected from 2003 to 2005. The stochastic frontier and technical inefficiency effects model is employed in the study, involving two input variables (i.e., nameplate generating capacity and implicit quantity of operating inputs), one output variable (i.e., electricity sent-out) and three explanatory variables (i.e., private ownership, plant age and year of observation). Our analysis finds that Malaysian electricity generating companies are relatively efficient in comparison to the selected international electricity generating companies, with an average TE score of 0.87 relative to an average TE score across all utilities of 0.79. Once again, the empirical results may be explained in a number of ways. First, one could argue that the superior performance of the Malaysian plants is perhaps a consequence of the use of second best data (i.e., aggregate input variables and approximate price deflators). Alternatively, the empirical results may be reliable and hence some commentators may argue that further market reform is not necessary. However, this conclusion should not be drawn until better quality data can be organised and additional analyses conducted. Finally, this thesis has made a number of key contributions. We have reviewed lessons learned from international electricity market reforms, and discussed their relevance to the case of Malaysia. We have compiled three detailed data sets and conducted three comprehensive empirical analyses of performance, from which we have (tentatively) concluded that private entry in the 1990’s did not result in improved TFP growth rates; that privately-owned generation plants are more efficient than publicly-owned plants; and that Malaysian power plants have above average performance relative to international comparators. However, a key theme in all three analyses is the fact that we have been forced to use a number of second-best measures in our data set, and hence that further research is needed in order for our important empirical results to be verified and then incorporated into the ongoing policy discussions in Malaysia regarding possible future reforms of the industry.