Regulation's Role in Stifling Innovation
From the KP archives
I’m traveling for work this week, so I’d like to revisit a speech I gave at an electricity interoperability conference called Grid Interop in December 2010. Most of the audience were engineers, so I focused my remarks on the economics of innovation, the economics of regulation, and the relationship between regulation and innovation. The topic is even more relevant today in electricity, and it’s also relevant to the growing discussion over AI regulation and the extent to which prescriptive AI regulation would stifle innovation (and would favor incumbents by raising their rivals’ costs).
We are brought together here by a shared interest in exploring, and expanding, interoperability and innovation in the electricity industry. Many of you come from a technology and/or engineering background. In contrast, I am an economist who studies regulation and its effects on innovation and new technology adoption. Today I’d like to use that perspective, along with some of the lessons I learned during the 5 enjoyable years during which I had the honor of serving on the GridWise Architecture Council, to explore the basic economic foundations of the regulatory environment in which technology and interoperability policy and investment decisions are being made. By “basic” I will abstract from the administrative, rate case, RFP weeds that we encounter regularly, and will instead focus on the economic concepts and incentives inherent in regulation.
In my remarks I’ll certainly say things you won’t like, or that you may not have considered, but my goal is to provide some provocative food for thought for you about the effects of economic regulation on your devices, interfaces, and business models. There’s also a certain degree of comfort in the known status quo, with large-scale utility procurement. But the economic institutions (by which I mean rules) and incentives that have created that status quo over a century will continue to limit innovation and growth unless they evolve.
The fundamental economic reason I make that argument is that traditional economic regulation is based on cost recovery, not on expected value creation, and therefore does a poor job of “standing in for the market” as it is (incorrectly) supposed to do in theory.
For the past century, economic regulation in the electricity industry has been based on a static economic model called natural monopoly theory. This model describes an industry with decreasing long-run average cost and a given demand for a given product (the combination of which creates economies of scale over the relevant range of demand, AKA subadditivity of costs for you econ geeks). The model suggests that the natural equilibrating tendency in a rivalrous market is to end up charging a price that does not cover the substantial fixed costs required to provide the infrastructure for the delivery of the good or service. Therefore, the designers of this model argued starting in the 1890s, economic regulation that erects an entry barrier, grants a monopoly, and stipulates that the firm’s profits will be a cost-plus rate of return will lead to the provision of that given product to those given consumers at the lowest feasible cost. Thus economic regulation defines market boundaries, and product and service boundaries (and quality), for a static demand curve. It accommodates demand growth simply by scaling up permitted infrastructure investment. Just in this high-level description of the economic theory underpinning our existing regulation, you can see how many crucial, static assumptions are made – many of which are increasingly violated in our vibrant, dynamic society.
Whether it’s enshrined in the legislation giving the regulatory agency its mission or in the deeply-embedded Populist culture and history of regulation, traditional regulatory procedures focus regulators and the regulated on providing a narrowly-defined, generic, highly reliable service at the lowest possible long-term cost. As long as you’re in a static environment, the static model from which this theory and culture emanate will do a decent job of providing that generic service. We experienced that with the policy focus on electrification through the 20th century. That’s the context in which regulators have developed a norm and a culture of ignoring value creation — focusing narrowly on the provision of generic electricity service and scoping your efforts accordingly fits with that static world.
But regulatory models premised on cost recovery fail in a more dynamic context, with pervasive economic and technological change and Schumpeterian creative destruction. As Exhibit A illustrating how vibrant this technological dynamism is, I give you this: my Android phone [Note: that 2010 link is dead, so instead consider my Google Pixel and how it has changed in 13 years from that HTC phone-ed.]. That dynamism characterizes the economic and technological context of the early 21st century, and the reason that dynamism and creative destruction become so pervasive in human society is that they create value — value for consumers, variety for consumers, product differentiation for consumers, and value for the risk-taking and opportunity-seeking entrepreneurs who risk private capital to create that value. That value arises through the breakdown of product definitions, service definitions, and market boundaries. Traditional economic regulation is designed precisely to prevent such value-creating innovation and evolution.
Remember that natural monopoly regulation forecloses entry by imposing a legal entry barrier. Thus as technology evolves, the appropriateness of the natural monopoly designation cannot be put to a market test. Furthermore, this lack of a market test undermines any potential disruptive innovation, so efficiency gains through technological change will be incremental and small.
From the perspective of consumers, the regulatory compact was intended to protect consumers, particularly the residential consumers who are also voters. It has done so by creating a policy environment in which the sole value proposition that regulators recognize as being “in the public interest” is one of keeping prices (rates) low and stable. In other words, a consequence of the regulatory compact is that the concept of consumer benefit in the electricity industry is narrower than in any other industry. By regulatory fiat, consumers only benefit from low, stable rates. The regulatory compact rigidifies the definition of consumer benefit, despite the fact that in many other industries, technological change and economic growth have created consumer benefit through innovation, new products and services, and product differentiation. As long as the regulatory compact retains the idea that consumer benefit in this industry derives only from low, stable rates, it prevents electricity consumers from having access to potentially valuable new products and services associated with electricity consumption, because it stifles innovation.
Innovation in digital technology has contributed to bringing us more choice as consumers. We now take as given that we will have choices when making consumption decisions. Choice is almost ubiquitous in our society, except for some network infrastructure industries, including electric power. When we consume electric power service we do not expect the same product differentiation, the same anticipation of the undiscovered desires of the consumer. Why not? Do we continue to believe that electric power service is merely the provision of a commodity that is an input into the consumption of other goods and services? If so, is that perception correct, or has technology changed the nature of the potential set of services surrounding electric power that firms can sell to consumers? Furthermore, if we consider service reliability, do we really believe that all consumers want to pay enough never to have an outage, but that they are willing to accept any amount of outage from weather?
These traditional economic regulatory institutions persist to this day; they have evolved in some ways, but in scope and in procedure they are essentially the regulatory institutions designed in the early 20th century to deal with the economic challenges arising from the economies of scale and subadditivity of cost. The original social compact persists, and the obligation to serve is a significant driver in utility incentives and regulatory policy.
One problem with this institutional persistence is that these regulatory institutions are premised on a static theory of human and firm behavior, but in reality the environment in which these institutions operate is dynamic. Particularly in the past 50 years, with the advent of digital technology, the growth of economic activity as more transactions shift out of firms and into markets, globalization, and fundamental demographic and labor market changes, the environment in which the electricity industry operates is extremely dynamic and fluid. This observation is also true for consumers, who have become accustomed to having variety and choice in consumer products and services, to having substantial (some would say excessive) information available to them, and to being able to control much of their personal environment and many of their transactions. Whether as a producer or consumer (or both), our lives have become more transactional, and much of our activity has shifted from taking place within firms and families to taking place through transactions mediated by market processes. This shift has increased wealth and human welfare significantly.
Regulatory institutions are not adaptive and generally do not deal well with change, particularly the effects of technological change. They arose in and are conditioned on a specific social-technological context in the early 20th century that no longer exists. The policy challenges and objectives present then (such as ubiquitous electrification and the imposition of control on monopoly profits) have receded and been reprioritized along with new policy challenges: providing incentives to innovate and adopt new technologies, providing the type of customer choice and decentralized decision-making capabilities that consumers experience in nearly all other aspects of their lives, managing the environmental consequences of electricity consumption, and the overarching objective of maintaining reliable service in the face of consistent, pervasive change.
Why have regulatory institutions not adapted to changes in the underlying, fundamental environment in which they operate? In part the answer to that question is because institutional change is incremental and takes time. However, there are still cases in which institutional change in response to disruptive fundamental change (such as, for example, sudden transformative technological change) has been beneficial. Some examples include the interstate highway system, the Internet, and the GPS satellite network; none of these systems, though, arose under the control of a regulatory agency. Indeed, there would even be some benefit from having institutional change anticipate future technological and economic change, but such foresight is difficult if not impossible. Thus the objective should be to have regulatory institutions that are adaptable, that can adapt to unknown and changing conditions. Adaptability means shifting the focus of regulatory institutions from specific details (like rate case reviews and top-down proscriptive decision-making) toward transparent principles, rules, and frameworks focused on consumer information and promoting competition.
In brief, traditional economic regulation is incompatible with economic dynamism, with technological change, with innovation, and ultimately incompatible with widespread consumer well-being because of the enormous extent to which traditional economic regulation stifles experimentation. The really valuable function that market processes provide is this ability for consumers and producers to experiment. Traditional economic regulation is almost reflexively anti-experimentation, and that reflex is the source of lost value creation opportunities from smart grid technologies.
The exciting work that many of you are doing opens up alternatives, particularly in complexity engineering and transactive energy. For the past century the dominant paradigm has been centralized economic and physical control based on natural monopoly theory and power systems engineering. The ideas presented and synthesized here compose a different paradigm – decentralized economic and physical coordination through contracts, transactions, price signals, and integrated intertemporal wholesale and retail markets. Digital communication technology, and its increasing pervasiveness and affordability, make this decentralized coordination possible. This decentralized coordination differs from the “distributed control” concept that power systems engineers often invoke; distributed control in that context means using distributed technology to enhance centralized control of a system. What I mean by decentralized coordination is a paradigm in which distributed agents themselves control part of the system, and in aggregate their actions produce order: emergent order. The GridWise Olympic Peninsula Project is an example of how we can achieve reliability and enhanced value through decentralized coordination. Technology makes this emergent order feasible, but the institutions, the rules governing the interaction of agents in the system, contribute substantially to whether or not order can emerge from this decentralized coordination process.
A historical example illustrates why I think these points are important. In the medieval period, China was one of the most forward-looking, open, technologically creative and vibrant societies in the world. Chinese inventions became the foundation of many important technologies, machines, and industries. Yet by 1600, China’s backwardness was obvious to all observers; China had closed herself off from knowledge, had become technologically stagnant. Western Europe and then the young United States surged ahead of China in technology, in economic productivity, in per-capita income, and in living standards for most of the population (China’s elite, of course, continued to enjoy luxury). Economic historians credit this stagnation and worsening of living standards for most of the Chinese population to conscious technocratic policy decisions in China to look inward (growing through population growth and increasing intensification of agriculture), to be backward-looking, and to make strong top-down rules based on status quo bias. Writ large, the dynamic driving the stultification of China had at its core many of the same policy drivers and incentives as we seen in play in electricity regulation in the 21st century.
If we are to achieve the widespread economic potential that innovation creates in the electricity industry, and if we are to create the conditions for further innovation, then regulation must evolve
Away from the limiting focus on cost recovery
Away from mandated entry barriers, especially in demonstrably competitive sectors like generation and retail
Toward an approach that’s more focused on consumer information and the promotion of competition as a policy objective
Toward a regulatory model that can adapt to unknown and changing conditions, and will therefore facilitate resilience, reliability, innovation, and value creation.
This is pure poetry. ❤️ “rigidifies”
This is a really great piece, thank you! But why not conclude by advocating for the removal of the regulatory oversight though instead of merely improving it?