Skip to content

Is Distributed Generation the Answer to Regulatory Dysfunction?

One delightful aspect of teaching an MBA course in energy and environmental markets is getting together with my former students as they pursue careers in the industries I study.  I learn so much about the latest trends and ideas in these markets, and they frequently challenge the way I have been seeing the world.

This happened recently when I had coffee with a former student whom I will refer to as “Pat”.  Pat has worked for a successful alternative energy company and done well, but s/he is ready to think about new paths.  Like many cleantech mavens, Pat is excited about distributed generation (DG), particularly with improving storage technologies.  Pat explained to me a potential business model s/he has been exploring with rooftop solar photovoltaic (PV) panels and on-site storage.

As I’ve written in a previous blog, I’m skeptical that rooftop solar is the most cost effective way to utilize the fabulous breakthroughs in PV technology.  I proceeded to lay out my argument, addressing each of the claims for distributed generation, even though I know Pat is a regular reader of the Energy Institute blog and had surely heard my views before.

But Pat was a star student and continues to be one of the most insightful people I know in the business.  So I was not surprised, but still unsettled, when Pat put on the table an argument for DG that I hadn’t heard before, or maybe Pat just presented it much more clearly so that I finally actually got it.

DysfunctionalUtilities1Here’s my dramatic (if you are an energy geek) re-creation of what Pat said: “Yes, Severin, in theory grid scale generation and delivery of renewable electricity generation is probably more cost-effective.  And, yes, there are some fixed cost of distribution systems that utilities are recovering through volumetric charges, which drives up the retail price and gives an inefficient incentive to install DG.  And, yes, California’s extreme increasing-block residential price schedules mean many households are paying more than 30 cents per kWh for much of their consumption, way above cost.”

“But,” Pat continued with growing enthusiasm, “California’s investor-owned utilities currently charge average residential rates in the 21 to 24 cent range –more than 50% above national average–and the utilities themselves are forecasting those numbers will rise in the coming years.  [Actually those are average rates among customers who aren’t on the low-income tariff.  More on that below. –SB]  I don’t know if rates are so high because of utility incompetence, a dysfunctional regulatory process, or some other reason, but it’s not my job to figure it out.  In any other industry, if a company’s prices are too high we rely on pressure from competition to reign them in.  Why should electricity be any different?”

Pat concluded with, “Severin, ever since I took your class many years ago you’ve been saying that California has high electricity rates in part to pay for the mistakes of the past.  But those ‘mistakes’ keep happening and keep driving up our rates.  At some point, aren’t those ongoing mistakes just part of a broken regulatory process? DG is the competition that will either force repairs in the process or will replace it.”

DysfunctionalUtilities2Pat’s argument isn’t entirely general; there are plenty of states — and even some municipal utilities in California — with rates that rooftop solar can’t touch.  And, there’s not much evidence nationally or internationally that competition introduced by deregulating retail electricity markets has significantly lowered rates.   Plus, it’s worth remembering that most residential customers don’t have a single-family home with a south-facing roof and no shading to put solar panels on, so most of us have to get all our electricity from the grid.

Nonetheless, Pat raises an important point.  Before proponents of high fixed charges and special fees for solar customers get too far down that road, they need to confront the fact that average residential electricity rates in California (and New York, and some other locations where DG is gaining the most traction) are out of line with the rest of the country.

I’ve been asking around about the high, and rising, average residential rates in California, and been surprised at the lack of clarity for the reasons. This seems like a central question of rooftop solar policy (as opposed to rooftop solar politics).  If the rates really reflect high costs of providing electricity, Pat and other DG supporters have a more compelling case that they are providing efficient competition.  On the other hand, if they are driven by other regulatory or legislative policy objectives, then we have to recognize that funding them in this way may encourage inefficient DG installation.

Put differently, is DG the answer to regulatory dysfunction, or is it just regulatory arbitrage? By regulatory arbitrage, I mean taking advantage of the structure of pricing or other utility obligations by pursuing strategies that reap private rewards through cost shifts to other ratepayers.

The simplest cause of regulatory arbitrage is the fact that electricity prices are well above the marginal cost of delivering a kilowatt-hour to the customer in California and many other states. In California, this is in part because of the regulator’s longtime resistance to fixed monthly charges, and in part because of the increasing-block price structure that leaves many customers today paying over 30 cents for their incremental kilowatt-hour.

In addition, the many programs that policymakers have decided to finance through electricity charges also invite regulatory arbitrage. For instance, significant parts of electricity bills in California and many other states pay for energy efficiency programs, early investments in renewable technologies, and — especially large in California — reduced electricity rates for low-income customers. Among the three large investor-owned utilities in California about 30% of all residential customers are on low-income rates.  And, of course, for more than a decade, part of electricity rates in California have paid to subsidize rooftop solar, both directly through the California Solar Initiative (from 2007 to 2013) and indirectly through net metering policies.

If all of these programs were eliminated, would average residential rates among California’s IOUs still be well above national average?   Of course, there are other factors that a cost analysis has to account for, such as the mix of generation, the density of residential consumers and the average consumption per customer.

I think that answering this question is critical to making good energy policy in California.  But after asking a number of regulators, utilities and other policy analysts in the state, I have not turned up any studies that put together all the numbers one needs.

That wouldn’t be the complete answer to Pat’s argument. It has to be paired with a credible analysis of the value and costs DG brings to the grid. But next time I see Pat, I’m hoping to have a better response than “good question. I should write a blog about that.”


I’m still tweeting energy news and research articles @BorensteinS

Severin Borenstein View All

Severin Borenstein is Professor of the Graduate School in the Economic Analysis and Policy Group at the Haas School of Business and Faculty Director of the Energy Institute at Haas. He received his A.B. from U.C. Berkeley and Ph.D. in Economics from M.I.T. His research focuses on the economics of renewable energy, economic policies for reducing greenhouse gases, and alternative models of retail electricity pricing. Borenstein is also a research associate of the National Bureau of Economic Research in Cambridge, MA. He served on the Board of Governors of the California Power Exchange from 1997 to 2003. During 1999-2000, he was a member of the California Attorney General's Gasoline Price Task Force. In 2012-13, he served on the Emissions Market Assessment Committee, which advised the California Air Resources Board on the operation of California’s Cap and Trade market for greenhouse gases. In 2014, he was appointed to the California Energy Commission’s Petroleum Market Advisory Committee, which he chaired from 2015 until the Committee was dissolved in 2017. From 2015-2020, he served on the Advisory Council of the Bay Area Air Quality Management District. Since 2019, he has been a member of the Governing Board of the California Independent System Operator.

32 thoughts on “Is Distributed Generation the Answer to Regulatory Dysfunction? Leave a comment

  1. Steve Huntoon, you look only from 1997 forward but CA’s energy efficiency policies have been in place since the 1970s. Also, in your comparison you don’t consider what might have happened in CA if EE policies weren’t in place. CA average useage could have risen, but it didn’t. Also, CA is far from temperate all year ’round or throughout its extent. Yes it’s warmer than Wisconsin but we also have very significant AC electricity needs that aren’t present in colder climates.

  2. I know you’re talking about residential, but, to get an idea of how high these tariffs are, consider a real time price (RTP) tariff that adds the LMP to the volumetric charges from the A1 commercial non-generation tariff. I calculate that the deadweight loss from that tariff is about 8 to 20 times as large as the deadweight loss from the optimal flat rate tariff, the latter being the mean LMP plus externalized carbon cost. So taking historical prices as given, and looking only at energy costs, the size of the markups seems to be a much bigger story than the lack of real time pricing.

    The non-gen charges, which CCA customers pay, are about 8-12 times the externalized carbon cost (social cost of carbon minus permit price), and they are still less than the implied volumetric T&D component that PG&E customers pay (which you get by subtracting the load-weighted LMP from the actual tariff).

    This is following the analysis in Borenstein 2005, assuming approximately linear supply and demand curves. Assuming that the marginal carbon impact is constant, because its variance is low and I don’t have the necessary data yet, the deadweight loss from this hypothetical RTP tariff above is proportional to the square of . I assume the social cost of carbon is $38 per metric ton and take the carbon impact data from Callaway et al. The deadweight loss from the optimal flat tariff is proportional to LMP variance (or, LMP plus carbon impact variance). The former is wayyy bigger than the latter. Calculations to be released soon….

    Question: should such an analysis use load-weighting, as in Borenstein 2005, or just take simple averages and variances, as in Hogan 2014?

    Click to access RTPchap05.pdf

    Click to access Hogan_TOU_RTP_Newark_082314.pdf

    Click to access WP264.pdf

  3. I think there’s another dimension to the DG business model that shouldn’t be ignored, although I appreciate it wasn’t what your former student was interested in. I recently reviewed a large number of DG proposals that were effectively competing with high cost, and in the Australian market, not generally competitively sourced network extension or augmentation. Rural producers, unwilling or unable to pay high network costs for local network capacity augmentations or the extension of (reliable) networks to the connection point, are looking at high reliability behind-the-meter alternatives. A combination of solar, storage and diesel can bring high reliability to the customer; the networks’ pricing approach to extensions and augmentations provides very little encouragement for collective action, so individual customers are no worse off than they would otherwise be; and, installing your own solution is likely to be considerably faster than the response time for a network approved extension or augmentation. Of course, you still need to get approval from the network for your new DG, but properly sized that should present few difficulties, and, perversely from customers’ collective perspective, the customer installing the new equipment may benefit from a reduction in his/her network charges.

  4. This is a nice article that leaves out human behavior and other assets that on-site energy and battery banks offer. It reminds me of the head of AT&T in 1970 saying cellular was a whim in a country with 99.7% quality service and who would spend 15 times more for a unit of communication. Currently, consumers assume losses during power outages, equipment losses do to poor (and declining) electric power quality, and in some cases, exceedingly high electric sub-rates, expressed as peak & seasonal power rates, demand charges, and in deregulated states, ratchet or spot rates. In an internet-connected world, short losses or power or electric spikes scrambling microprocessors or burning out equipment can mean loss of business or impacting health and safety. In other words mimicing the value of cellular — convenience, safety, better data, and more individual control. The other problem I have with this paper is the assumption that centralized grids are somehow sacrosanct rather than the blending of the entire portfolio of high-value energy efficiency, energy storage, and on-site (or near-site) renewable and distributed generation. I think that supposition is nonsense, and have seen no modern peer-reviewed papers proving that. – Scott Sklar, Adjunct Professor, The George Washington University and Chairman, RE&EE Advisory Committee, US Department of Commerce

  5. I was struck by your question: “If all of these [energy efficiency / solar / low-income] programs were eliminated, would average residential rates among California’s IOUs still be well above national average? ”

    I’m surprised there isn’t any kind of average residential bill breakdown available… I’m guessing you’re referring to something like this ( but for California. I believe you can find something on your PG&E utility bill – mine includes a breakdown of the total bill into Generation, Transmission Distribution, Public purpose programs etc. so this must exist somewhere in a more general statewide format.
    The public purpose programs portion worked out at about 7% of my bill so based on this very scientific sample size of 1 I don’t think removing these would necessarily get rid of the big differential with the rest of the country, or dramatically alter my incentives on DG. It may be that there are other policies that are not being funded through this public purpose programs charge though.

    Specifically on your mention of CARE (now a $1.2 billion/yr program), I was actually looking at these figures recently as the data on what the CARE surcharge is for non-participants is published by the CPUC. You can get the data for all the IOUs, but I looked up just PG&E and found:
    – For residential customers the CARE surcharge is generally about 2-4% of their bill, which is $4-5 on elec and $1 on gas.
    – For commercial customers its pretty similar at around 3-6%, which is about $45 on elec and $7 on gas.
    – For industrials it’s around a 9-12% increase on their bill, which is about $9000 on elec and $6000 on gas.

    These certainly aren’t negligible, but again, is it enough to explain the big differential with other states – probably not. Interestingly, the much higher CARE surcharges being paid by industrials compared to residential maybe indicates that focusing on DG for residential customers is a bit too narrow. After all, it may be larger customers that have the biggest incentive to reduce the amounts they pay into these public programs by looking to DG.

    All this aside though, I am totally with you that a proper empirical look at what is driving the divergence with other states (public programs or not) would be a welcome piece of analysis!

  6. At the core of this discussion is what share of risk should utility shareholders bear? Currently they take on little–almost all of their mistakes, going back to Diablo and SONGS, are passed on entirely to ratepayers. (In the case of Diablo, we are paying double.) When the CPUC tries to move toward a “market” solution, they try to steer around the issue that the IOUs need to take on more risk with market mechanisms. Otherwise the incentives are too blunted. Florio’s recent scoping ruling on DER planning appears to make the same mistake. It’s time to move the other direction.

    Residential rates would not decrease much by dropping the direct public goods charges. The legacy contract costs and overinvestment in T&D are the main culprits in high rates (having worked on GRCs for many years.) Note that residential customers bear only about 40% of the subsidy to CARE customers–the rest come from the other customer groups.

    So I agree that solar provides strong competitive pressure, not only to the IOUs but also to regulators. That’s why Picker et al have made hostile noises about CCAs and rooftop solar. Regulators will lose their leverage over delivering policy goodies to key stakeholders who rely on the current centralized generation system. California does not have a strong track record for making good cost-effective planning decisions. Let’s allow consumers to make those decisions themselves.

    What’s surprising is that the payback on solar for even low-level consumers who have bills around a $100 a month is relatively short. That seems to show that the choice is cost effective over a wide array of users, regardless of the rates being charged.

    Jim R: note that California has 4-tier rates, not 2-tier, which is why they are so inefficient and distorting. Everyone sees different incremental rates.

    • Richard,
      California is right for having escalatiing rates but for the wrong reason. If they had the right reason, they would only need two rates. Rates should be set for improved (social) efficiency. Poverty alleviation should be left to other instruments.

      • Jim, I agree up to a point, but the reality is that we have multiple tiers for political reasons here. I think those may be disappearing as we move to TOU rates, but we need to talk within the context of howe things exist, not how we wished that they existed.

      • Richard,
        If we have to have an imperative about efficiency vs. reality, I’ll go w/ the Dixit rule. Start w/ 1st-best analysis to assess what might improve efficiency. Use 2nd-best analysis to account for information and enforcement problems. Finally, apply 3rd-best analysis to incorporate political feasibility. If equity is the political goal of escalating tiers, one can rationalize 4 tiers. The problem is that this is the “wrong” reason according to the Musgrave-Kaplow-Shavell principle that there are more cost effective instruments for reducing poverty. Channeling Wicksell, efficiency-based pricing reform should therefore be packaged w/ enhanced use of poverty reduction instruments.
        P.S. Had to reply to myself, since they don’t let you reply to a reply.

        • Jim, as an economist i absolutely agree with your general analytic approach. The problem in California is that they have politically and legally ruled out that approach so far. The four tier system looks like it will be collapsed to lesser tiers, but we are stuck with the current structure and the escalating tiers lead to ridiculous differentials. If economists want to have an influence on the political debate, and not just publish irrelevant articles in irrelevant journals, they have to start from the current institutional arrangements and understanding what is politically feasible. “Coulda woulda” statements are not effective.

      • Richard,
        I’m sure you know this already: “Practical men who believe themselves to be quite exempt from any intellectual influence, are usually the slaves of some defunct economist.”
        — J.M.K.

  7. What is the reason for fixed charges in the first place? Two-part tariff? Power is not a decreasing cost industry. The efficient generation mix increases from low-cost (e.g. hydro) to higher social cost sources as total generation increases. Economies in transmission have likewise been exhausted (Vernon Smith 1979) and may be negative, e.g. due to increasing right-of-way costs. But isn’t infrastructure fixed in the short-run thereby warranting fixed charges? That ignores the fact that, e.g. w/ fixed transmission capacity, short run marginal costs are increasing in transmission, due to line resistance and depreciation. (Similar argument w/ fixed generation assets.) Of course, fixed costs could be justified by uncertainty, but they could just as easily be negative as positive.
    That leaves the policy errors of the past that SB talks about. Yes, this sort of cost recovery would be better handled by fixed rather than volumetric charges, but why burden electricity consumers any more than you would charge only bank customers for a financial bailout? It’s clearly a violation of horizontal equity. The real explanation must lie in the realm of political economy and the fact that other California taxes are already high.
    If there are indeed increasing marginal costs of power expansion, there is nothing wrong in principle w/ escalating tiered rates. The inefficiency derives from facing different consumers with different marginal rates. You only need two rates, and the goal is to have as many customers as possible facing the second (marginal cost) rate. Finally, using escalating power rates as an instrument of poverty alleviation ignores the Musgrave-Kaplow-Shavell principle of using the most cost effective instrument first.

  8. Ha! Rooftop solar and distributed generation IS competition for traditional electric utilities. The question that should be asked is how utilities should respond to competition. Distributed generation is also how wise consumers should be reacting NOW to protect themselves from future rate increases driven by: (1) accelerating environmental compliance-related costs and (2) shift of common costs re-allocated from more price elastic commercial and industrial consumers to less elastic residential consumers.

    When competition came to telecom, the reaction of monopoly telecoms was: (1) act to prevent market entry (did not work); (2) penalize competitors (did not work); (3) attempt to keep telecoms “whole” by shifting the embedded subsidies around (e.g., from CPE, to long distance, to access charges, etc) or creating new subsidy mechanisms (did not work); and (4) create disconnected “bundles” of services the utility hoped its consumers would want (e.g., sateliite TV and local telephone service), thus, stay with the former monopoly telecom provider. Few of the traditional telecoms could cope with competition and many went from being profitable monopolies to acquisition targets where the acquirer hoped to be sucessful through economies of scale.

    In my neck of the woods – Colorado – electric rates are not so high as California, but when my electric utility closed a single coal fired power plant my monthly rate went from around $100/month to $175-$225. I’m sure more is to come.

    When electric rates increase for commercial and industrial consumers, such customers can and do react by installing alternative distributed generation facilities (e.g., solar panels on the rooftop of the local WalMart) or, simply leaving the jurisdiction and relocating a facility in a lower cost state. The common costs formerly covered by commercial and industrial consumers that respond to higher prices are typically shifted to residential consumers in the largely arbitrary cost allocation process of a rate case. Steve Huntoon’s comment suggests that this may be happening already.

%d bloggers like this: