Restructuring Revisited Part 1: Competition in Electricity Distribution Systems.

Author:Burger, Scott P.

    The emergence of distributed energy resources (DERs) (1) and digital technologies are creating new options for the delivery of electricity services and the potential for more affordable and resilient power systems. However, these developments are also placing new strains on electric power industry structures that were established in a time of static distribution networks and relatively inelastic demand. DERs and digital technologies dramatically expand the number of potential investors in and operators of power system infrastructure, creating a new set of actors capable of competing for the provision of electricity services.

    During the wave of restructuring that swept through the electricity industry in the 1980s, 1990s, and 2000s, regulators grappled with questions about the value of enabling or encouraging competition in the generation and (to a lesser extent) retail sectors. This spawned a set of questions over how to assign the roles of transmission system ownership and operation, generation ownership, wholesale energy market operation, and retailing to power system actors in order to ensure efficient pricing and the development of an affordable mix of transmission and generation assets in the short and long run (European Commission, 2009; FERC, 1999; Joskow, 1996; Joskow & Schmalensee, 1983).

    Today, the emergence of DERs is spurring regulators to engage in analogous debates over which actors should perform which roles within the distribution system. One of the primary objectives of these debates is maintaining or enhancing competition in the horizontal segments of the power sector (e.g., generation and retailing) where it exists, while potentially fostering more competitive provision of "non-wires" or operational alternatives to investment in conventional network assets. Existing industry structures need to be revisited once again, to guarantee that they do adequately achieve these goals.

    This is the first paper in a two-part series that explores the implications of decentralization and digitization for competition and coordination in electricity distribution systems. Part 1 (this paper) analyzes the economic characteristics of the distribution-level roles required for efficient planning and operation of the power system and the implications of these characteristics for industry structure, competition, market development, and economic efficiency. Part 2 focuses on price and market structures needed to enable coordination in planning, investment, and operation of distribution systems and connected devices.

    This paper begins by defining and reviewing the core activities and economic characteristics of four key industry roles: distribution network ownership (DNO); distribution system operation (DSO); DER ownership and operation; and aggregation of demand and DERs. We analyze the potential for separating distribution system operation and ownership through a transaction cost lens. We then discuss the monopoly nature of the DNO and DSO roles, and consider the implications of their monopoly characteristics on adjacent roles that may be suitable for competition: DER ownership and aggregation. Throughout this document, our analysis focuses primarily on whether a given structure is likely to lead to the welfare maximizing (2) mix of network, generation, and demand-side resources in the short and the long run.

    This paper has implications for three questions that are currently being debated by regulators and policy makers globally:

  2. Should distribution system operations be separated from distribution network ownership in order to ensure the neutrality of the DSO role?

  3. Should DNOs be allowed to own and operate DERs, or should DER ownership be left exclusively to competitive actors?

  4. Does the emergence of DERs necessitate a reconsideration of the role of competition in the provision of aggregation services such as retailing?

    We find that separating distribution system operations and network ownership creates significant transaction costs and is likely to be less efficient than a system in which the DNO and DSO are an integrated, single entity. In addition, we find that there is limited evidence to support DNO or DSO ownership of DERs in regions in which the DNO is unbundled from generation. Based on the best available data, competitive providers are likely to provide lower cost DERs relative to monopoly providers, and enabling monopoly ownership of such assets may threaten the functioning of competitive markets. Finally, DSOs are becoming responsible for a much greater degree of system operation and planning and may increasingly procure services from a variety of distributed network users and aggregators, making effective separation of monopoly and competitive activities imperative, just as it is in transmission network and bulk power system operations.


    The bulk of the literature on electric power sector reform and the ensuing industry structure--developed during the 1980s, 1990s, and early 2000s--focuses on the use of transaction cost economics and comparative institutional analysis and centers around the assignment of responsibilities for bulk power generation, transmission system ownership and operation, and, to a lesser degree, retailing functions. The transaction cost economics framework states that the structure of the power sector--that is, the extent of vertical and horizontal integration--is the efficient outcome of attempts to coordinate the sector's many interlinked roles and to manage the complexities introduced by the network externalities (losses and congestions) that are pervasive in power systems (Joskow and Schmalensee 1983).

    Transaction cost economics, first articulated by Williamson (1971, 1975, 1979) dictates that firms face choices to perform certain activities or functions (e.g. production of a good) within the firm or to contract with external parties for these activities or functions. Where contracting involves highly complex activities, is infrequent, involves durable and specific assets, faces large degrees of uncertainty over the value of these assets, and/ or where the quality of the assets is difficult for the contracting party to verify, firms will typically seek to integrate these activities into their operations. In this context, integration can increase efficiency by minimizing transaction costs. Vertical integration can also increase efficiency by preventing double marginalization (1) (Spengler, 1950) and enabling more efficient investment in infrastructure (Hart & Moore, 1990; Williamson, 1975). Thus, setting aside the potential for negative impacts on competition and market power, the efficient outcome under some conditions is vertical integration.

    Substantial empirical evidence validates these theories across a number of industries (Lafontaine & Slade, 2007) and within the power sector in particular (Joskow, 1985, 1987). Alternative theories of vertical integration based on evaluation of property rights (Hart & Moore, 1990) or moral hazard (Lafontaine & Slade, 2007) exist, but transaction cost analyses, such as those performed by Joskow and Schmalensee (1983), Joskow (1996), and Newbery (2002) have dominated the literature on vertical integration in the electricity sector and the introduction of competition in power generation and retail aggregation.

    However, vertical integration is not costless. It creates the possibility for the vertically integrated firm to foreclose (that is, raise the costs of or bar entry by) competitors (Rey & Tirole. 2007). This is especially true in the electric power industry, where certain horizontal segments of the sector (transmission and distribution) are natural monopolies and control physical access between suppliers and consumers. In addition, where vertically integrated firms act as monopolies, they do not face the incentives for efficiency created by competition. Regulators attempt to create incentives for efficiency, but regulation is never perfect and often fails to create effective incentives (Cicala, 2015; Laffont & Tirole, 1993). Furthermore, markets have been effective at creating competitive incentives for efficiency in generation (Borenstein & Bushnell, 2015; Fabrizio, Rose, & Wolfram, 2007; Joskow, 2006). Thus, when considering any given industry structure, we must consider the potential for both costs and benefits from various degrees of integration.

    Some recent work has applied this transaction cost framework to the distribution level. For example, van Werven and Scheepers (2005) argue that the emergence of DERs creates the need for more active distribution system operations, and that enabling such operations requires stronger unbundling of network operation and competitive service provision as well as better incentive regulation for distribution utilities. Bauknecht and Brunekreeft (2008) and Brunekreeft and Ehlers (2006), in contrast, argue that the benefits of improved coordination of investments in network infrastructure and DERs justify vertical integration. Eurelectric (2016) highlights the need for the DSO to remain neutral in future power systems. Brandstatt et al. (2016) highlight the benefits of independent access to data for enhancing competition and coordination in the power system.

    MIT (2016) reviews the lessons learned from restructuring conversations in the 1990s and 2000s and discusses the pros and cons of three potential options for distribution level institutional arrangements. The primary options discussed by MIT (2016) are: 1) a distribution system operator independent from the distribution network owner, 2) a combined distribution network owner and operator that is independent from all competitive activities, and 3) a vertically integrated and closely regulated utility. This paper builds upon this previous work by considering further evidence, the multitude of potential structures at the distribution level, and the potential efficiencies gained...

To continue reading