“Grid defection” and the regulated utility business model

The conversations about the “utility death spiral” to which I alluded in my recent post have included discussion of the potential for “grid defection”. Grid defection is an important phenomenon in any network industry — what if you use scarce resources to build a network that provides value for consumers, and then over time, with innovation and dynamism, what if they can find alternative ways of capturing that value (and/or more or different value)? Whether it’s a public transportation network, a wired telecommunications network, a water and sewer network, or a wired electricity distribution network, consumers can and do exit when they perceive the alternatives available to them as being more valuable than the network alternative. Of course, those four cases differ because of differences in transaction costs and regulatory institutions — making exit from a public transportation network illegal (i.e., making private transportation illegal) is much less likely, and less valuable, than making private water supply in a municipality illegal. But two of the common elements across these four infrastructure industries are interesting: the high fixed costs nature of the network infrastructure and the resulting economies of scale, and the potential for innovation and technological change to change the relative value of the network.

The first common element in network industries is the high fixed costs associated with constructing and maintaining the network, and the associated economies of scale typically found in such industries. This cost structure has long been the justification for either economic regulation or municipal supply in the industry — the cheapest per-unit way to provide large quantities is to have one provider and not to build duplicate networks, and to stipulate product quality and degrees of infrastructure redundancy to provide reliable service at the lowest feasible cost.

What does that entail? Cost-based regulation. Spreading those fixed costs out over as many consumers as possible to keep the product’s regulated price as low as feasible. If there are different consumers that can be categorized into different customer classes, and if for economic or political reasons the utility and/or the regulator have an incentive to keep prices low for one class (say, residential customers), then other types of consumers may bear a larger share of the fixed costs than they would if, for example, the fixed costs were allocated according to share of the volume of network use (this is called cross-subsidization). Cost-based regulation has been the typical regulatory approach in these industries, and cross-subsidization has been a characteristic of regulated rate structures. The classic reference for this analysis is Faulhaber American Economic Review (1975).

Both in theory and in practice these institutions can work as long as the technological environment is static. But the technological environment is anything but static; it has had periods of stability but has always been dynamic, the dynamism of which is the foundation of increased living standards over the past three centuries. Technological dynamism creates new alternatives to the existing network industry. We have seen this happen in the past two decades with mobile communications eroding the value of wired communications at a rapid rate, and that history animates the concern in electricity that distributed generation will make the distribution network less valuable and will disintermediate the regulated distribution utility, the wires owner, which relies on the distribution transaction for its revenue. It also traditionally relies on the ability to cross-subsidize across different types of customers, by charging different portions of that fixed costs to different types of customers, and that’s a pricing practice that mobile telephony also made obsolete in the communications market.

Alternatives to the network grid may have higher value to consumers in their estimation (never forget that value is subjective), and they may be willing to pay more to achieve that value. This is why most of us now pay more per month for communications services than we did pre-1984 in our monthly phone bill. As customers leave the traditional network to capture that value, though, those network fixed costs are now spread over fewer network customers. That’s the Achilles heel of cost-based regulation. And that’s a big part of what drives the “death spiral” concern — if customers increasingly self-generate and leave the network, who will pay the fixed costs? This question has traditionally been the justification for regulators approving utility standby charges, so that if a customer self-generates and has a failure, that customer can connect to the grid and get electricity. Set those rates too high, and distributed generation’s economic value falls; set those rates too low, and the distribution utility may not cover the incremental costs of serving that customer. That range can be large.

This is not a new conversation in the industry or among policy makers and academics. In fact, here’s a 2003 Electricity Journal article arguing against standby charges by friend-of-KP Sean Casten, who works in recycled energy and combined heat and power (CHP). In 2002 I presented a paper at the International Association of Energy Economics annual meetings in which I argued that distributed generation and storage would make the distribution network contestable, and after the Northeast blackout in 2003 Reason released a version of the paper as a policy study. One typical static argument for a single, regulated wires network is to eliminate costly duplication of infrastructure in the presence of economies of scale. But my argument is dynamic: innovation and technological change that competes with the wires network need not be duplicative wires, and DG+storage is an example of innovation that makes a wires network contestable.

Another older conversation that is new again was the DISCO of the Future Forum, hosted over a year or so in 2001-2002 by the Center for the Advancement of Energy Markets. I participated in this forum, in which industry, regulators, and researchers worked together to “game out” different scenarios for the distribution company business model in the context of competitive wholesale and retail markets. This 2002 Electric Light & Power article summarizes the effort and the ultimate report; note in particular this description of the forum from Jamie Wimberly, then-CAEM president (and now CEO of EcoAlign):

“The primary purpose of the forum was to thoroughly examine the issues and challenges facing distribution companies and to make consensus-based recommendations that work to ensure healthy companies and happy customers in the future,” he said. “There is no question much more needs to be discussed and debated, particularly the role of the regulated utility in the provision of new product offerings and services.”

Technological dynamism is starting to make the distribution network contestable. Now what?

The “utility death spiral”: The utility as a regulatory creation

Unless you follow the electricity industry you may not be aware of the past year’s discussion of the impending “utility death spiral”, ably summarized in this Clean Energy Group post:

There have been several reports out recently predicting that solar + storage systems will soon reach cost parity with grid-purchased electricity, thus presenting the first serious challenge to the centralized utility model.  Customers, the theory goes, will soon be able to cut the cord that has bound them to traditional utilities, opting instead to self-generate using cheap PV, with batteries to regulate the intermittent output and carry them through cloudy spells.  The plummeting cost of solar panels, plus the imminent increased production and decreased cost of electric vehicle batteries that can be used in stationary applications, have combined to create a technological perfect storm. As grid power costs rise and self-generation costs fall, a tipping point will arrive – within a decade, some analysts are predicting – at which time, it will become economically advantageous for millions of Americans to generate their own power.  The “death spiral” for utilities occurs because the more people self-generate, the more utilities will be forced to seek rate increases on a shrinking rate base… thus driving even more customers off the grid.

A January 2013 analysis from the Edison Electric Institute, Disruptive Challenges: Financial Implications and Strategic Responses to a Changing Retail Electric Business, precipitated this conversation. Focusing on the financial market implications for regulated utilities of distributed resources (DER) and technology-enabled demand-side management (an archaic term that I dislike intensely), or DSM, the report notes that:

The financial risks created by disruptive challenges include declining utility revenues, increasing costs, and lower profitability potential, particularly over the long term. As DER and DSM programs continue to capture “market share,” for example, utility revenues will be reduced. Adding the higher costs to integrate DER, increasing subsidies for DSM and direct metering of DER will result in the potential for a squeeze on profitability and, thus, credit metrics. While the regulatory process is expected to allow for recovery of lost revenues in future rate cases, tariff structures in most states call for non-DER customers to pay for (or absorb) lost revenues. As DER penetration increases, this is a cost recovery structure that will lead to political pressure to undo these cross subsidies and may result in utility stranded cost exposure.

I think the apocalyptic “death spiral” rhetoric is overblown and exaggerated, but this is a worthwhile, and perhaps overdue, conversation to have. As it has unfolded over the past year, though, I do think that some of the more essential questions on the topic are not being asked. Over the next few weeks I’m going to explore some of those questions, as I dive into a related new research project.

The theoretical argument for the possibility of death spiral is straightforward. The vertically-integrated, regulated distribution utility is a regulatory creation, intended to enable a financially sustainable business model for providing reliable basic electricity service to the largest possible number of customers for the least feasible cost, taking account of the economies of scale and scope resulting from the electro-mechanical generation and wires technologies implemented in the early 20th century. From a theoretical/benevolent social planner perspective, the objective is, given a market demand for a specific good/service, to minimize the total cost of providing that good/service subject to a zero economic profit constraint for the firm; this will lead to highest feasible output and total surplus combination (and lowest deadweight loss) consistent with the financial sustainability of the firm.

The regulatory mechanism for implementing this model to achieve this objective is to erect a legal entry barrier into the market for that specific good/service, and to assure the regulated monopolist cost recovery, including its opportunity cost of capital, otherwise known as rate-of-return regulation. In return, the regulated monopolist commits to serve all customers reliably through its vertically-integrated generation, transmission, distribution, and retail functions. The monopolist’s costs and opportunity cost of capital determine its revenue requirement, out of which we can derive flat, averaged retail prices that forecasts suggest will enable the monopolist to earn that amount of revenue.

That’s the regulatory model + business model that has existed with little substantive evolution since the early 20th century, and it did achieve the social policy objectives of the 20th century — widespread electrification and low, stable prices, which have enabled follow-on economic growth and well-distributed increased living standards. It’s a regulatory+business model, though, that is premised on a few things:

  1. Defining a market by defining the characteristics of the product/service sold in that market, in this case electricity with a particular physical (volts, amps, hertz) definition and a particular reliability level (paraphrasing Fred Kahn …)
  2. The economies of scale (those big central generators and big wires) and economies of scope (lower total cost when producing two or more products compared to producing those products separately) that exist due to large-scale electro-mechanical technologies
  3. The architectural implications of connecting large-scale electro-mechanical technologies together in a network via a set of centralized control nodes — technology -> architecture -> market environment, and in this case large-scale electro-mechanical technologies -> distributed wires network with centralized control points rather than distributed control points throughout the network, including the edge of the network (paraphrasing Larry Lessig …)
  4. The financial implications of having invested so many resources in long-lived physical assets to create that network and its control nodes — if demand is growing at a stable rate, and regulators can assure cost recovery, then the regulated monopolist can arrange financing for investments at attractive interest rates, as long as this arrangement is likely to be stable for the 30-to-40-year life of the assets

As long as those conditions are stable, regulatory cost recovery will sustain this business model. And that’s precisely the effect of smart grid technologies, distributed generation technologies, microgrid technologies — they violate one or more of those four premises, and can make it not just feasible, but actually beneficial for customers to change their behavior in ways that reduce the regulation-supported revenue of the regulated monopolist.

Digital technologies that enable greater consumer control and more choice of products and services break down the regulatory market boundaries that are required to regulate product quality. Generation innovations, from the combined-cycle gas turbine of the 1980s to small-scale Stirling engines, reduce the economies of scale that have driven the regulation of and investment in the industry for over a century. Wires networks with centralized control built to capitalize on those large-scale technologies may have less value in an environment with smaller-scale generation and digital, automated detection, response, and control. But those generation and wires assets are long-lived, and in a cost-recovery-based business model, have to be paid for even if they become the destruction in creative destruction. We saw that happen in the restructuring that occurred in the 1990s, with the liberalization of wholesale power markets and the unbundling of generation from the vertically-integrated monopolists in those states; part of the political bargain in restructuring was to compensate them for the “stranded costs” associated with having made those investments based on a regulatory commitment that they would receive cost recovery on them.

Thus the death spiral rhetoric, and the concern that the existing utility business model will not survive. But if my framing of the situation is accurate, then what we should be examining in more detail is the regulatory model, since the utility business model is itself a regulatory creation. This relationship between digital innovation (encompassing smart grid, distributed resources, and microgrids) and regulation is what I’m exploring. How should the regulatory model and the associated utility business model change in light of digital innovation?

The political economy of Uber’s multi-dimensional creative destruction

Over the past week it’s been hard to keep up with the news about Uber. Uber’s creative destruction is rapid, and occurring on multiple dimensions in different places. And while the focus right now is on Uber’s disruption in the shared transportation market, I suspect that more disruption will arise in other markets too.

Start with two facts from this Wired article from last week by Marcus Wohlsen: Uber has just completed a funding round that raised an additional $1.2 billion, and last week it announced lower UberX fares in San Francisco, New York, and Chicago (the Chicago reduction was not mentioned in the article, but I am an Uber Chicago customer, so I received a notification of it). This second fact is interesting, especially once one digs in a little deeper:

With not just success but survival on the line, Uber has even more incentive to expand as rapidly as possible. If it gets big enough quickly enough, the political price could become too high for any elected official who tries to pull Uber to the curb.

Yesterday, Uber announced it was lowering UberX fares by 20 percent in New York City, claiming the cuts would make its cheapest service cheaper than a regular yellow taxi. That follows a 25 percent decrease in the San Francisco Bay Areaannounced last week, and a similar drop in Los Angeles UberX prices revealed earlier last month. The company says UberX drivers in California (though apparently not in New York) will still get paid their standard 80 percent portion of what the fare would have been before the discount. As Forbes‘ Ellen Huet points out, the arrangement means a San Francisco ride that once cost $15 will now cost passengers $11.25, but the driver still gets paid $12.

So one thing they’re doing with their cash is essentially topping off payments to drivers while lowering prices to customers for the UberX service. Note that Uber is a multi-service firm, with rides at different quality/price combinations. I think Wohlsen’s Wired argument is right, and that they are pursuing a strategy of “grow the base quickly”, even if it means that the UberX prices are loss leaders for now (while their other service prices remain unchanged). In a recent (highly recommended!) EconTalk podcast, Russ Roberts and Mike Munger also make this point.

This “grow the base” strategy is common in tech industries, and we’ve seen it repeatedly over the past 15 years with Amazon and others. But, as Wohlsen notes, this strategy has an additional benefit of making regulatory inertia and status quo protection more costly. The more popular Uber becomes with more people, the harder it will be for existing taxi interests to succeed in shutting them down.

The ease, the transparency, the convenience, the lower transaction costs, the ability to see and submit driver ratings, the consumer assessment of whether Uber’s reputation and driver certification provides him/her with enough expectation of safety — all of these are things that consumers can now assess for themselves, without a regulator’s judgment substitution for their own judgment. The technology, the business model, and the reputation mechanism diminish the public safety justification for taxi regulation. Creative destruction and freedom to innovate are the core of improvements in living standards. But the regulated taxi industry, having paid for medallions with the expectation of perpetual entry barriers, are seeing the value of the government-created entry barrier wither, and are lobbying to stem the losses in the value of their medallions. Note here the similarity between this situation and the one in the 1990s when regulated electric utilities argued, largely successfully, that they should be compensated for “stranded costs” when they were required to divest their generation capacity at lower prices due to the anticipation of competitive wholesale markets. One consequence of regulation is the expectation of the right to a profitable business model, an expectation that flies in the face of economic growth and dynamic change.

Another move that I think represents a political compromise while giving Uber a PR opportunity was last week’s agreement with the New York Attorney General to cap “surge pricing” during citywide emergencies, a policy that Uber appears to be extending nationally. As Megan McArdle notes, this does indeed make economists sad, since Uber’s surge pricing is a wonderful example of how dynamic pricing induces more drivers to supply rides when demand is high, rather than leaving potential passengers with fewer taxis in the face of a fixed, regulated price.

Sadly, no one else loves surge pricing as much as economists do. Instead of getting all excited about the subtle, elegant machinery of price discovery, people get all outraged about “price gouging.” No matter how earnestly economists and their fellow travelers explain that this is irrational madness — that price gouging actually makes everyone better off by ensuring greater supply and allocating the supply to (approximately) those with the greatest demand — the rest of the country continues to view marking up generators after a hurricane, or similar maneuvers, as a pretty serious moral crime.

Back in April Mike wrote here about how likely this was to happen in NY, and in commenting on the agreement with the NY AG last week, Regulation editor Peter Van Doren gave a great shout-out to Mike’s lead article in the Spring 2011 issue on price gouging regulations and their ethical and welfare effects.

Even though the surge pricing cap during emergencies is economically harmful but politically predictable (in Megan’s words), I think the real effects of Uber will transcend the shared ride market. It’s a flexible piece of software — an app, a menu of contracts with drivers and riders, transparency, a reputation mechanism. Much as Amazon started by disrupting the retail book market and then expanded because of the flexibility of its software, I expect Uber to do something similar, in some form.

Building, and commercializing, a better nuclear reactor

A couple of years ago, I was transfixed by the research from Leslie Dewan and Mark Massie highlighted in their TedX video on the future of nuclear power.


A recent IEEE Spectrum article highlights what Dewan and Massie have been up to since then, which is founding a startup called Transatomic Power in partnership with investor Russ Wilcox. The description of the reactor from the article indicates its potential benefits:

The design they came up with is a variant on the molten salt reactors first demonstrated in the 1950s. This type of reactor uses fuel dissolved in a liquid salt at a temperature of around 650 °C instead of the solid fuel rods found in today’s conventional reactors. Improving on the 1950s design, Dewan and Massie’s reactor could run on spent nuclear fuel, thus reducing the industry’s nuclear waste problem. What’s more, Dewan says, their reactor would be “walk-away safe,” a key selling point in a post-Fukushima world. “If you don’t have electric power, or if you don’t have any operators on site, the reactor will just coast to a stop, and the salt will freeze solid in the course of a few hours,” she says.

The article goes on to discuss raising funds for lab experiments and a subsequent demonstration project, and it ends on a skeptical note, with an indication that existing industrial nuclear manufacturers in the US and Europe are unlikely to be interested in commercializing such an advanced reactor technology. Perhaps the best prospects for such a technology are in Asia.

Another thing I found striking in reading this article, and that I find in general when reading about advanced nuclear reactor technology, is how dismissive some people are of such innovation — why not go for thorium, or why even bother with this when the “real” answer is to harness solar power for nuclear fission? Such criticisms of innovations like this are misguided, and show a misunderstanding of both the economics of innovation and the process of innovation itself. One of the clear benefits of this innovation is its use of a known, proven reactor technology in a novel way and using spent fuel rod waste as fuel. This incremental “killing two birds with one stone” approach may be an economical approach to generating clean electricity, reducing waste, and filling a technology gap while more basic science research continues on other generation technologies.

Arguing that nuclear is a waste of time is the equivalent of a “swing for the fences” energy innovation strategy. Transatomic’s reactor represents a “get guys on base” energy innovation strategy. We certainly should do basic research and swing for the fences, but that’s no substitute for the incremental benefits of getting new technologies on base that create value in multiple energy and environmental dimensions.

Ben Powell on drought and water pricing

Ben Powell at Texas Tech has an essay on water scarcity at Huffington Post in which he channels David Zetland:

But water shortages in Lubbock and elsewhere are not meteorological phenomena. The shortages are a man-made result of bad economic policy.

Droughts make water scarcer, but by themselves they cannot cause shortages. To have a shortage and a risk of depletion, a resource must be mispriced.

With the freedom to choose, consumers can demonstrate whether it’s worth the cost to them to water their lawn an extra day or hose dust off of their house. Realistic pricing also incentivizes them to take account of water’s scarcity when they consume it in ways that aren’t currently prohibited. Have your long shower if you want . . . but pay the real price of it instead of the current subsidized rate.

Of course Ben is correct in his analysis and his policy recommendation, although I would nuance it with David’s “some for free, pay for more” to address some of the income distribution/regressivity aspects of municipal water pricing. Water is almost universally mispriced and wasted, exacerbating the distress and economic costs of drought.

The spin on wind, or, an example of bullshit in the field of energy policy

The Wall Street Journal recently opined against President Obama’s nominee for Federal Energy Regulatory Commission chairman, Norman Bay, and in the process took a modest swipe at subsidies for wind energy.

The context here is Bay’s action while leading FERC’s enforcement division, and in particular his prosecution of electric power market participants who manage to run afoul of FERC’s vague definition for market manipulation even though their trading behavior complied with all laws, regulations, and market rules.

So here the WSJ‘s editorial board pokes a little at subsidized wind in the process of making a point about reckless prosecutions:

As a thought experiment, consider the production tax credit for wind energy. In certain places at certain times, the subsidy is lucrative enough that wind generators make bids at negative prices: Instead of selling their product, they pay the market to drive prices below zero or “buy” electricity that would otherwise go unsold to qualify for the credit.

That strategy harms unsubsidized energy sources, distorts competition and may be an offense against taxpayers. But it isn’t a crime in the conventional legal sense because wind outfits are merely exploiting the subsidy in the open. The rational solution would be to end the subsidies that create negative bids, not to indict the wind farms. But for Mr. Bay, the same logic doesn’t apply to FERC.

The first quoted paragraph seems descriptive of reality and doesn’t cast wind energy in any negative light. The second quoted paragraph suggests the subsidy harms unsubsidized competitors, also plainly true, and that it “distorts competition” and “may be an offense against taxpayers.” These last two characterizations also strike me as fair descriptions of current public policy, and perhaps as mildly negative in tone.

Of course folks at the wind industry’s lobby shop are eager to challenge any little perceived slight, so the AWEA’s Michael Goggin sent a letter to the editor:

Your editorial “Electric Prosecutor Acid Test” (May 19) ignores wind energy’s real consumer benefits by mentioning the red herring of negative electricity prices. Negative prices are extremely rare and are usually highly localized in remote areas where they have little to no impact on other power plants, are caused by inflexible nuclear power plants much of the time, and are being eliminated as long-needed grid upgrades are completed.

Wind energy’s real impact is saving consumers money by displacing more expensive forms of energy, which is precisely why utilities bought wind in the first place. This impact is entirely market-driven, occurs with or without the tax credit, and applies to all low-fuel-cost sources of energy, including nuclear.

The tax relief provided to wind energy more than pays for itself by enabling economic development that generates additional tax revenue and represents a small fraction of the cumulative incentives given to other energy sources.

Michael Goggin
American Wind Energy Association
Washington, DC

Let’s just say I’ll believe the “impact is entirely market-driven” when someone produces a convincing study that shows the exact same wind energy capacity build-out would have happened over the last 20 years in the absence of the U.S. federal Production Tax Credit and state renewable energy purchase mandates. Without the tax credit, the wind energy industry likely would be (I’m guessing) less than one-tenth of its current size and without a big tax credit wouldn’t be the target of much public policy debate.

Of course, without much public policy debate, the wind energy industry wouldn’t need to hire so many lobbyists. Hence the AWEA’s urge to jump on any perceived slight, stir the pot, and keep debate going.

MORE on the lobbying against the Bay nomination. See also this WSJ op-ed.


Court says no to FERC’s negawatt payment rule

Jeremy Jacobs and Hannah Northey at Greenwire report “Appeals court throws out FERC’s demand-response order“:

A federal appeals court today threw out a high-profile Federal Energy Regulatory Commission order that provided incentives for electricity users to consume less power, a practice dubbed demand response.

In a divided ruling, the U.S. Court of Appeals for the District of Columbia Circuit struck a blow to the Obama administration’s energy efficiency efforts, vacating a 2011 FERC order requiring grid operators to pay customers and demand-response providers the market value of unused electricity.

Among environmentalists this demand-response enabled “unused electricity” is sometimes described as negawatts. FERC’s rule required FERC-regulated wholesale electric power markets to pay demand-response providers the full market price of electricity. It is, of course, economic nonsense pursued in the effort to boost demand response programs in FERC-regulated markets.

The court held that FERC significantly overstepped the commission’s authority under the Federal Power Act.

The Federal Power Act assigns most regulatory authority over retail electricity prices to the states, and the court said FERC’s demand response pricing rule interfered with state regulators’ authority.

Personally, I would have dinged FERC’s rule for economic stupidity, but maybe that isn’t the court’s job. Actually, I did ding the FERC’s rule for its economic stupidity. I was one of twenty economists joining in a amicus brief in the case arguing that the FERC pricing rule didn’t make sense. The court’s decision gave our brief a nod:

Although we need not delve now into the dispute among experts, see, e.g., Br. of Leading Economists as Amicus Curiae in Support of Pet’rs, the potential windfall  to demand response resources seems troubling, and the Commissioner’s concerns are certainly valid.  Indeed, “overcompensation cannot be just and reasonable,” Order 745-A, 2011 WL 6523756, at *38 (Moeller, dissenting), and the Commission has not adequately explained how their system results in just compensation.

But if this negawatt-market price idea survives the appeals court rejection and takes off in the energy policy area, I have the following idea: I’d really like a Tesla automobile, but the current price indicates that Teslas are in high demand so I’m going to not buy one today. Okay, now who is going to pay me $90,000 for the nega-Tesla I just made?