Christian Growitsch, Gernot Müller, Marcus Stronzik
Unbundling of Transmission System Operators in the gas sector - economic theory and empirical evidence
No. 308 / May 2008
Having investigated the electricity and gas sector, the European Commission was quite unsatisfied with the market outcomes. The current unbundling regime of network operators was identified as one major obstacle to a well functioning market environment.Therefore, in their third legislative package the Commission argues for an ownership unbundling (OU) of network operators in the energy sectors as the farthest-reaching regulatory regime of vertical disintegration. The intention of this paper is to contribute to this current discussion by empirically analyzing the effects of OU.
Focusing on gas, theoretical papers on the effects of OU hardly exist. Considering studies on the electricity sector, they do not provide a clear picture regarding the benefits of this regulatory measure. While these papers mostly find advantages with respect to end user prices, efficiency, quality and competitiveness as well as an easing of the regulatory process, the implications for investments are mixed. Furthermore, some papers point at higher transactions costs due to a higher complexity and stress the loss of economies of scope.
Empirical studies find that the overall liberalization of energy markets has generally led to a reduced level of end user prices. Likewise theoretical papers, they mostly focus on electricity. Concerning the regulatory impact of the various measures, they identify different factors crucial for the success of these reforms. These ambiguous results are due to different empirical approaches as well as to the time periods and countries taken into account. As they all use aggregated indictors in order to map the regulatory reforms, the effects of OU cannot be separated.
To the best of our knowledge this is the first paper which investigates the isolated effect of unbundling on end user prices. In doing so, we construct a panel out of 20 OECD countries spanning a time interval of 18 years. We control for the oil price, the GDP, structural parameters of the gas sector like energy supplied and indigenous production as well as for other regulatory variables like ownership structure, degree of market opening and third party access. Using advanced panel approaches like the fixed effects estimator with robust standard errors and a dynamic estimator which corrects for the least square dummy variable estimates by approximating the small sample bias, we find no significant effect of OU on prices for households. Instead, the variable for market opening shows significant influence in the static as well as in the dynamic setting. Against the background of the current political discussions the results indicate that regulatory measures should concentrate more on the overall market opening instead of taking the vertical disintegration a step further.
[only a german version available]
Discussion Paper is available for download.
- WIK_Diskussionsbeitrag_Nr_308_01.pdf330 Ki