Risks and Regulation, Part 2

Lynne Kiesling

A couple of days ago I outlined some remarks I gave at a regulation conference organized by the Istituto Bruno Leoni, and I asked what some of the risks are associated with electricity regulation. In the comments on that post, Ed Reid and David Zetland touched on what I think are two important risks arising from regulation — and not just in electricity:


You have written here on several occasions regarding smart meters, smart grid, variable rates, etc. Many of the programs which have attempted to evaluate the potential of these approaches have failed because regulators refused to expose the consumer participants in the test program to real time rates or established such short test program durations that many of the investment-intensive consumer response options offered no potential to be recovered from savings.

The smart grid is viewed by many as essential to a renewable energy future. However, the reluctance of regulators to permit construction and testing of an attractive consumer value proposition over a test period long enough to encourage permanent change, might “kill the goose” before it has the opportunity to lay any eggs.


The biggest risk is regulating uncertainty as if it’s risk, i.e., pretending one knows the probability and damages of an event and then attaching regulations to those guesses.

Their comments illustrate two points I made in my talk: regulation is prone to the problems of Bastiat’s unseen (we see the costs and benefits we experience, but not the ones that we fail to realize by choosing to foreclose retail dynamic pricing, bundling, and product differentiation), and in regulation we have a serious epistemological hubris problem because regulators make decisions based on regulating uncertainty as if it’s risk that have large impacts and are difficult to reverse. Add these to the risks I discussed before — reliability, bankruptcy/investment, environmental, fuel price volatility, and foregone value opportunities due to lack of innovation (which overlaps with Ed’s point) — and the risks we face from perpetuating our current regulatory system are large and diverse. Here are some of the issues I raised about them at the conference:

Set of risks: These risks interact, and regulatory attempts to reduce or eliminate one usually increase another or introduce a new one. For example, the focus on reliability and financial stability focus has led to static market definition associated with legal entry barriers, the commodification of the electricity product, and little innovation (just put iron in the ground, customers don’t want that digital stuff). Such a traditional regulatory focus reduces the incentive and ability of firms in the electricity industry to discover and commercialize new technologies, products, and services with lower emission footprints, so the focus on reliability and bankruptcy risk reduction contributes to increasing environmental risk. There are other examples, but you get the idea.

Innovation changes everything: A centralized or top-down regulatory approach to these risks ignores consumer preferences, the heterogeneity of consumer preferences, and the fact that consumer preferences can change over time as technology changes. Take, for example, consumer end-use digital technologies for energy management (programmable communicating thermostats, web portal, etc.). Such devices empower consumers to program their energy consumption/willingness to pay preferences into their homes and devices, to automate responses to price signals, to choose the level of reliability that they want/are willing to pay for (even by room!), and even to choose how their electricity is generated (green/grey mix). Of course the canonical example of research into this value proposition from innovation is the GridWise Olympic Peninsula project that I’ve discussed here many times before.

With such technology, the static, standalone commodified retail electricity market is an anachronism. With such technology and other distribution-related smart grid technologies, product differentiation (dynamic pricing, green power, differentiated reliability) is possible, bundling with home entertainment or home security or home health monitoring is possible, and both consumers and competitive retail energy service providers can benefit … and in the process they can create ways to manage this set of risks, ways that are almost certain to differ from what would be imposed top-down using the traditional regulatory techniques of the electro-mechanical era.

We can’t eliminate risk, but we can price it through markets: Risk is not a bad thing, not an aspect of life than should, or can, be eliminated. But it can be priced – and should be priced through the methods I just described, which are decentralized market processes for entrepreneurs to discover and create products and services that benefit consumers. Some of those entrepreneurs will fail, and that’s as it should be, because well-being and thriving human societies only arise from systems where individuals can profit OR lose from their endeavors.

Resiliency: Another benefit of such decentralized systems (this applies to financial regulation too, even more so than energy) is resilience – their distributed nature means that a failure of any one entity does not materially affect the system as a whole. Two examples here are (a) unexpected physical failure of a generator and its effects, which can be mitigated through distributed end-use response to the price increase that would naturally accompany such an unexpected outage; and (b) exit of a retail firm from the market, where establishing a set of market rules for customer service in the event of exit minimizes customer disruption while still offering them alternatives from competitors. Pluralism and decentralization in complex systems makes them more resilient, which itself contributes to changing the risk portfolio of the system. When unavoidable failures happen, pluralism and decentralized market processes allow for systems to fail gracefully.

Imperfection: Even if we made all of those (beneficial IMO) changes to make retail electricity markets more competitive, more innovative, and more able to price risk, the imperfection of human systems is inescapable. In the world I sketched out, consumers may not choose options that change their energy consumption in ways that align with “optimal carbon objectives”, if we could know and define such a thing. Incomplete property rights that are costly to define are going to continue to be the core economic cause of environmental issues, including greenhouse gases.

One overarching issue in regulation of all the forms we’re discussing is control vs. choice – both history and human nature put us on a path of desiring control over our circumstances. But history also shows us that such control is an illusion, particularly in complex systems, and paradoxically, is it through choice in decentralized systems that we can achieve the kind of stability, reliability, and thriving that are the stated objectives of regulation.

Thus I conclude that if we try to use regulation to reduce risks or eliminate them entirely, we do so at a substantial cost. Our best approach to risks involving energy and the environment is to have an institutional framework that allows experimentation and trial & error, because only experimentation will enable us to discover the institutional and technological innovations that we can use to create more resilient human systems for pricing and managing the complex set of risks we face in this area.

3 thoughts on “Risks and Regulation, Part 2

  1. In the previous article, you criticized command and control regulation as “inefficient because it cannot reflect, as Hayek said “individual knowledge of time and place””

    Here, you say “Even if we made all of those (beneficial IMO) changes to make retail electricity markets more competitive, more innovative, and more able to price risk, the imperfection of human systems is inescapable.”

    Command and Control regulation is always a bad thing in efficient markets, but, ” the imperfection of human systems is inescapable.” Energy markets are far from efficient. Hence, there are many instances where command and control regulation leads to greater market efficiency. For example, set top boxes are massive energy wasters because of split incentives: the people who pay for the energy are not the same people who chose the set top boxes: they are provided by the cable company. The cable company has an incentive to purchase the cheapest possible set top box, because they will not be paying for the electricity it uses. The consumer does not have a choice, and muct use the box provided by the cable company, and probably does not know anything about its electricity usage anyway.

    In this, and many similar cases, the most efficient solution is not the (broken) market solution, but command and control regulation: government mandated standards for set top box efficiency.

    We have to be careful not to apply the efficient markets paradigm to markets that are far from efficient.

  2. One important aspect you touch upon is that regulation can not – and should not – eliminate risk. It may reduce, or even eliminate, a specific risk, and by so doing it usually increase some other risk (or some other undesirable feature). For example, if you want to reduce the climate risk, you may outlaw carbon-based energy sources. That would certainly save climate (under the assumption that climate change is man-made), but that would also, for example, increase energy poverty and increase the probability and severity of shortages, shocks, etc. In a sense, this happens in the market, as well: whenever you – as a power producer or consumer – make a technological choice, or decide that reliability is more important than affordability (or viceversa), you are trading a risk for another. The difference, though, is that, by and large, individual choices affect just the individual who makes them. She or he can learn from the others, and others may learn from her or him. That would allow, all else being equal, for more trials, more errors, and eventually more progress and more innovation. When, instead, regulators make a choice, it is almost by definition under a “one size fits all” kind of presumption – and very often what size is picked responds not just to the best of our (imperfect) scientific understanding, but also to (much more perfect) knowledge of rent seekers about what would best accomodate their own interests.
    Even Tom’s example of set top boxes power efficiency is not very persuasive: it relies on the un-proven and un-demonstrable assumption that a more costly, more efficient set top box is (socially) more desirable than a cheaper, less energy-efficient one. The only possible answer is that there saving energy is SO important that it must come at any cost – including the cost of having less people with a set top box (if extra costs are passed onto consumers) or having less investments in cable TV (if extra costs fall upon the cable company), or a combination thereof. I don’t really see any “objective” reason for that. The only possible, plausible, and questionable reason might be that we don’t fully pay for externalities deriving from energy consumption: but IF that is the case, there much more efficient ways to make people pay for the externalities they cause, than regulating set top boxes.

  3. Carlo,
    My set top box example is much stronger than you imply. This is because the cost of a more efficient set-top box is only a couple of bucks, and will pay itself back to the consumer in a month or two. Unless we apply a discount rate well over 100%, this would bring a net economic gain to both the consumer and society.

    The cable company will pass on the higher costs of set-top boxes to cable customers, who will end up being economically better off, because the savings in energy costs will be much greater than the extra subscription fees.

    Inefficient markets are inherently more risky than efficient ones, because if true risks are reflected in prices, the market will seek less risky solutions. When regulation makes markets more efficient, it reduces risk. We just have to be very careful about the regulations we choose, as many regulations add inefficiency rather than vice versa.

Comments are closed.