Vertical Separation of Transmission Control and Regional Production Efficiency in the Electricity Industry.

• Author: Chu, Yin

1. INTRODUCTION

   Economists generally believe that promoting competitive markets serves to enhance efficiency and welfare, evidence of which has been found by a series of empirical analyses across a wide spectrum of industries. (1) In this spirit, one of the most recent market restructuring transformations in the U.S. occurred in the electricity industry. Until the mid-1990s, the U.S. power industry was largely comprised of vertically integrated utilities in the chain of generation, transmission and distribution, operating as local natural monopolies. Regulated utilities were compensated under the "rate-of-return" principle to cover their costs plus a fair return. Agency models indicate that under the regulation structure, firms would deviate from cost-minimization behaviors as regulators setting the prices are asymmetrically informed (Laffont and Tirole, 1993). Integrated utilities also have incentives to over-utilize their own facilities and provide discriminatory transmission service to non-integrated wholesale competitors to protect their sales for revenue compensation. Given these concerns, market restructuring activities have been enacted in several states since the mid-1990s, which provoked a considerable body of economic studies evaluating the impacts on the performance of the electricity industry. (2)

   This study analyzes the welfare implications of one specific aspect of the market restructuring process in the U.S. electricity industry. Typically, restructuring may consist of the following aspects: (1) separating the transmission function from vertically integrated utilities, (2) allowing wholesale pricing typically via a centralized bidding mechanism, (3) divesting generation assets from retailers, and (4) allowing customers to switch retailers. In order to evaluate the impacts of restructuring for policy recommendations, researchers must disentangle these channels, which is generally a difficult task. (3) The Southwest Power Pool (SPP) market, however, provides a venue to separate them because this market only experienced market restructuring in the transmission component. Taking advantage of this unique market, I investigate the potential efficiency gains brought about by the vertical separation of transmission control in the U.S. electric power industry. In practice, this is achieved by establishing an organized wholesale market intermediated by a Regional Transmission Operator (RTO) or Independent System Operator (ISO), which takes over the transmission control from previously integrated utilities. (4)

   The necessity of separating the transmission sector from others is largely grounded on the principle that an electricity market functions effectively only under the condition of non-discriminatory transmission access. Given the network nature of the electric power industry, transmission access is vital since competing power producers must rely on it to schedule and dispatch their generating units. Before deregulation, a vertically integrated utility firm not only participated in the downstream product market (i.e., electricity generation), but also supplied an essential upstream input (i.e., transmission access). Previous theoretical literature have demonstrated that a vertically integrated firm has incentives to practice non-price discrimination against downstream rivals and raise their costs through quality degradation of the upstream input (Weisman, 1995; Economides, 1998; Beard, Kaserman and Mayo, 2001). (5) Particularly, Beard, Kaserman and Mayo (2001) show that an effective upstream input price regulation would in turn create a perverse incentive for the integrated firm to downgrade the quality of the upstream input. Examples of such "sabotage" form of discrimination were documented in network industries (i.e., energy, telecommunications, etc.) in EU (Hoffler and Kranz, 2011). (6) This implies that facing assess tariff regulation, a vertically integrated utility firm would potentially discriminate against non-integrated competing generators through quality degradation of the transmission services. Despite the extensive theoretical literature and great policy relevance, there have been relatively few empirical analyses on the efficiency impacts of the vertical separation. This study represents an intellectual endeavor to fill this gap.

   In this paper, I look into the impact of the divestiture of transmission control from vertically integrated utility producers on regional production efficiency. The question of interest is: can such separation lead to better allocation of production resources and increase the probability of lowcost generating units being dispatched over high-cost ones? If it is the case that vertically integrated utility producers engage in transmission discrimination and over-utilize their own generating assets, outside lower-cost options would be potentially under-utilized. This would lead to an inefficient allocation of the regional production resources. With transmission control vertically separated and handed over to an impartial RTO, the possibility of discriminatory transmission access is removed. Under enhanced wholesale competition, under-utilized cost-efficient plants would have incentives to produce more. Intuitively, this would improve regional production efficiency.

   Following Douglas (2006), I measure regional production efficiency through the sensitivity of unit utilization with respect to average costs. The implicit logic is that regions where production resources are allocated more efficiently should rely more on low-cost generating units, rather than over-utilizing high-cost ones. Accordingly, the utilization of generators in such an environment should be more responsive to their own average costs. Under this logic, I employ the difference-in-difference strategy and compare the average cost sensitivity of unit utilization in the SPP market with a control market, where no market restructuring activities ever took place.

   I utilize an 8-year monthly panel (2001-2008) of detailed micro-data at the generating unit level and choose the establishment of the RTO-monitored wholesale market in the SPP as the treatment event. Based on robust empirical results, I fail to find significant market wide evidence of improvement in regional production efficiency associated with the vertical separation of transmission control. However, looking into subgroups of generators, I find mixed evidence of cost savings via reallocation of production resources. On the one hand, coal units are dispatched more efficiently after the restructuring, indicated by an increase in the cost sensitivity of unit utilization by 8%-11%. On the other hand, such efficiency gains are not found for two types of gas units with different combustion technologies and cost efficiency. The results may be attributed to a lack of market-oriented tools for information revelation: gas units generally operate in peak-load periods when transmission constraints are more likely to bind and matching the supply and demand is generally more challenging.

   My study contributes to the literature in several aspects. First, by extending the analysis to a distinct organized wholesale market, this study adds to the literature by disentangling and assessing the effect of deregulation on one specific component of the electricity sector: the electricity transmission network. Earlier studies on market restructuring fail to disentangle it from other efficiency-enhancing channels, such as privatization of production assets, establishment of centralized wholesale market platforms and so on. Identifying the impact of each efficiency enhancing channel separately is vital for policy recommendations on the optimal design of restructuring "packages". Owning to high implementation costs and incidents of market crises (e.g., the California debacle during 2000-2001), deregulation has been seriously challenged in the U.S. and heavily debated in other parts of world. Therefore, better understandings on how to design the deregulation "package" is even more emphasized. This study suggests that although the vertical separation of transmission control may be appealing for policy makers who favor a minor level of restructuring, the efficiency improvement associated with it is also limited. Restructuring needs to go beyond it to obtain efficiency gains of sufficient scale.

   Second, by looking into dispatch efficiency at the market level, this paper complements the large literature on the deregulation in the U.S. power industry, the vast majority of which examine plant- or unit-level operating efficiency (Fabrizio, Rose and Wolfram, 2007; Zhang, 2007; Davis and Wolfram, 2012; Craig and Savage, 2013; Hausman, 2014; Cicala, 2015; Chan, Fell, Lange and Li, 2017). Two exceptions are Douglas (2006) and Mansur and White (2012). Douglas (2006) examines market restructuring in late 1990s in the eastern regions, yet without disentangling various channels of efficiency improvement. Mansur and White (2012) compare two typical wholesale market mechanisms, decentralized bilateral trading vesus centralized auction, and find that a market structure change from the former to the latter substantially improved the overall market efficiency.

   Third, this study also adds fuel to discussions on the cost-and-benefit comparison between vertical integration and separation of the network infrastructure, which receives considerable policy debates in energy sectors. For instance, given inquiry on the role of vertically integrated incumbents in the energy sector, the EU commission adopted a package of energy proposals in September 2007, one of which is the separation of transmission from production and supply in the electricity and gas sectors. By evaluating the impact on the optimal allocation of regional production resources, this study represents one of the few empirical studies in the literature of vertical separation.

   The remainder of the paper proceeds as follows...