Vertical Integration and Wholesale Power Markets

Should the same organization operate both transmission grids and wholesale markets?

On September 4, 1882, Thomas Edison opened the Pearl Street Station in lower Manhattan. Pearl Street embodied several of Edison's inventions: the first commercially viable incandescent light bulb, the "Jumbo" dynamo (mechanical steam generator), and the safety fuses and other devices required to provide safe electric lighting service. But it also embodied another original contribution of Edison to electrification: electric service as an integrated system. Edison envisioned and designed a deliberately vertically-integrated system, where every component from the Jumbo dynamo to the lighting fixtures and bulbs inside the home were designed to work together and to be sold together as a single, integrated service. Little did he realize how his vision would persist, and how the dynamics of digital and generation technology innovation would put pressure on his system in ways that we are grappling with in 2023.

In many ways the system Edison refined at Pearl Street was simpler than the one that would ultimately win the War of Systems a decade later. In 1882 Pearl Street Station could serve 1,200 lights in one square mile. Edison designed the system using direct current, which limited the distance the current could travel before losses grew too large. He planned to duplicate this setup, copy-paste style, to provide electric lighting service to more people. This was not a system that could achieve the economies of scale and scope that we've come to associate with electricity. Over the next decade the vision and inventions of his former colleague Nikola Tesla would dramatically reimagine the possible technologies, architecture, and uses of electric service, leading to the Current War/War of Systems between Edison's direct current model and the alternating current (AC) model of Tesla and George Westinghouse, which enabled higher-voltage transmission of current over longer distances with less loss of energy as it traveled.

AC systems proliferated and were the dominant model by the turn of the 20th century. By 1920 the new architecture was firmly in place: large "central station" generation situated at a distance from population centers delivering electricity to passive consumers over high-voltage transmission and low-voltage distribution wires networks, with consumption measured at the end-use premises by a mechanical meter that was read at the end of the month. Despite all of the changes in the technologies and the architecture, the system model remained Edison's vertically-integrated vision. One could make a technological determinism argument for this integration – with large electro-mechanical technologies and an essential requirement for real-time balancing in an era without digital switches or sensors, delivering safe, reliable, and affordable electric service necessitated single ownership of the entire vertical supply chain.

In institutional and transaction cost economics, we model vertical integration as a consequence of transaction costs that make it costly for/prevent Generator A transacting with Retail Provider B, who arranges for transportation through a separate transaction with Wires Company C. In the case of electricity, the technology characteristics and system architecture essentially create transaction costs that make it more valuable, profitable, and efficient to organize production through a vertically-integrated firm rather than three separate firms (or four, if the transmission wires company and distribution wires company are separate). These transaction costs are not static, though, and one big factor changing transaction costs over time is innovation – new markets emerge for new products and services, new methods of organizing production, new technologies that make exchange through markets and contracts feasible that weren't before.

Transaction costs aren't the only driver of vertical integration in electricity; the other is the aforementioned economies of scale and scope that reinforce the economics of organizing production in large vertically-integrated firms (which if you wanted to call indirect transaction costs I wouldn't quibble with). This driver of vertical integration started to recede in the 1980s, when the invention of the combined-cycle gas turbine generator changed the monolithic economies of scale in generation. Generation technology heterogeneity eroded the economies of scale in generation, and combining that erosion with the falling transaction costs from computerized financial market platforms starting around the same time enabled the evolution of wholesale power markets in the 1990s.

This evolution required some legislative and regulatory institutional change, because the Federal Power Act of 1935 had pretty firmly entrenched the vertically-integrated regulated utility business model while also requiring regulatory oversight of "just and reasonable" pricing in wholesale regional power pools that were starting to form for emergency backup purposes. The Energy Policy Act of 1992 removed barriers to the formation of independent power producers and a competitive generation model, enabling the unbundling of the generation portion of the vertically-integrated supply chain and the formation of competitive wholesale power markets. Over the 1990s these markets formed regionally, building on pre-existing regional power pools that in some regions (e.g., Mid-Atlantic) had existed since the 1920s and on some development of non-utility generators enabled by the Public Utilities Regulatory Policy Act (PURPA) in 1978.

The other regulatory change that established wholesale power market structure was Federal Energy Regulatory Commission (FERC) Order 888 (and 889) in 1996, with the goal "to remove impediments to competition in the wholesale bulk power marketplace and to bring more efficient, lower cost power to the Nation's electricity consumers". Order 888 required transmission owning utilities that engage in interstate commerce to have "posted offer" nondiscriminatory open-access pricing of transmission; FERC subsequently issued Order 2000 in 1999 to specify the characteristics and functions of Regional Transmission Organizations that would operate these regional markets.

The concern here illustrates challenges in the dynamics of vertical integration: when a regulated utility used to own both generation and transmission, and now it either has unbundled its generation or has "firewalled" it in another business unit that now offers generation in wholesale markets, how should the regulatory institutional framework structure the institutions to ensure that the transmission owner won't offer preferential pricing to their affiliated or recently-affiliated generators? It's an admittedly strong set of regulations intended to reduce entry barriers that could stymie competition, reduce innovation, and make consumers worse off.

Regulators making these institutional changes paid attention to market power of both generators and transmission owners, which is why FERC required transmission owners to adopt open-access tariffs. But they didn't pay attention to the odd organizational structure that would ensue and the governance implications of having regulated transmission-owning utilities as partners in RTOs that would operate wholesale power markets that were designed to be competitive. RTO participation separates transmission ownership and operation, which can lead to misaligned investment incentives (particularly since transmission is still a regulated function). Another governance issue is the extent to which decision-making reflects the interests of incumbent generators as a group of powerful market participants, with rules that accordingly reflect their interests (Green & Laffont 1979).

One way to think about the RTO organizational structure is that it's a transmission operating joint venture that's vertically integrated with a market platform, but that a powerful set of market participants influence the rules by which the JV operates and the market design. Consider the analogy to a financial futures market for, say, cattle. If that market were structured and governed like an RTO, the railroads and trucking companies would be part of the joint venture and the large meat processors would have an outsized voice in the financial market rules. The analogy has limits because cattle futures and delivery do not require real-time balancing like electricity does, but it’s a useful thought experiment.

Since the late 1990s, regional wholesale power markets have grown and now serve two-thirds of the US (and have also grown internationally, which I won't discuss here). The Western states are in the process of establishing an RTO and a wholesale market, and a 2020 Vibrant Clean Energy study suggests both economic and environmental benefits would arise from a southeastern wholesale market. Paul Joskow summarized the benefits of competitive wholesale markets in his 2008 Energy Journal article:

The benefits are to be realized by relying on competitive wholesale markets for power to provide better incentives for controlling construction and operating costs of new and existing generating capacity, to encourage innovation in power supply technologies, to provide incentives for network operators to provide appropriate levels of service quality, and to shift the risks of technology choice, construction cost and operating “mistakes” to suppliers and away from consumers. (p. 11)

Steve Cicala's recent article, Imperfect Markets versus Imperfect Regulation in U.S. Electricity Generation (2022) estimates the cost reductions attributable to wholesale power markets compared to cost estimates if plants instead continued to operate under vertical integration. One big difference between the two is how scarce transmission capacity is allocated:

Centralized wholesale electricity markets (“market dispatch”) integrate dispatch operations into an auction for electricity. In day-ahead auctions, for example, generators submit bids to produce electricity, and only those below the price needed to meet projected demand are called on to operate. These auctions incorporate feasibility constraints, so calling on higher-priced units to operate due to transmission congestion allows for the direct revelation of the cost of shortcomings in the transmission system." (p. 414)

This use of market dispatch reduced costs compared to conventional dispatch:

[M]arket-based dispatch has caused a 16 percent reduction in out-of-merit costs, while increasing gains from trade by 55 percent—a reduction in production costs of between $3 and $5 billion per year. These savings are worth roughly 5 percent of the total variable cost of generating electricity in market areas. (p. 438)

Wholesale power markets have lately also been instrumental in enabling innovation and investment in clean energy technologies, an unanticipated benefit from the perspective of the 1990s. A recent articulation of these benefits comes from Bill Massey, a FERC commissioner 1993-2003 and former head of the COMPETE Coalition that advocated for continued development and innovation to improve wholesale market competition.

But wholesale markets are not without problems. From the outset RTOs have had conflicted governance, with the wedge between transmission ownership and operation and much of the decision-making power in many RTOs resting with generators. This power dynamic influences decisions they make about how best to achieve objectives like reliability, resource adequacy, and long-term planning, and since many decisions are made through a disproportionately generation-focused lens, they overlook many of the benefits that can arise from non-generation approaches to those objectives. These approaches have emerged through technological and organizational change over the past two decades. On this point I think Massey is unfortunately too optimistic when he says

Organized regional markets also provide the level playing field for clean energy to compete without discrimination. That’s because the regional market operator is independent of market participants. It owns no generation or transmission resources; in other words, it has no “skin in the game,” and thus no incentive to favor or shun any particular resource, a problem that can exist outside the organized markets where clean energy providers are sometimes shut out from competing on fair terms.

His comparison is likely accurate, though, as it's true that non-utility resources face entry barriers in vertically-integrated states, but organized wholesale markets are not without discrimination issues.

Another problem is that over the past decade, "out of market" state environmental policy decisions have diluted the most important feature of any market: meaningful, informative price signals that emerge from the process of price discovery in markets with rules that are designed to elicit bids and offers that reflect true underlying marginal benefit and marginal cost. And the reliability, resilience, and planning challenges that ERCOT saw in winter storm Uri in 2021 and MISO and PJM saw in Elliott in 2022 illustrate the challenges of operating wholesale markets in very complex, interdependent systems. Both of these topics are worthy of further analysis at another time.

I don't have an answer, but I agree with Ray Gifford and Matt Larson that these challenges are important to analyze and to consider what future power markets look like. I am not persuaded by their conclusion that we should focus on standalone "transco" transmission utilities as a way to remove transmission ownership and operation decisions from wholesale market operation decisions. I haven't decided yet whether they are asking the wrong question or just asking the question the wrong way, but perhaps the question should be instead framed as unbundling the market platform from the RTO, rather than unbundling transmission.

When Edison designed and built his vertically integrated system, he could not have conceived of digital market platforms and how they could combine with other new technologies to make a less integrated system preferable. Part of the conversation I think we need to have, in research and in policy analysis, is identifying better institutional designs for enhancing the integrity of market design and meaningful, informative prices signals while enabling low entry barriers, separate from but coordinated with transmission operation.