Skip to content

Rate design wars are the sound of utilities taking residential PV seriously

Imagine walking into your supermarket with a bag of zucchini from your garden and saying that you’d like to trade them straight up for an equal quantity of zucchini next month.

The store manager would explain that they aren’t in the business of making wholesale purchases at such small scale, and that when they do make wholesale purchases it is at a much lower price than the retail price at which they sell.


Swapping zucchini today for zucchini tomorrow

You can, of course, eat the zucchini you grow, the manager might say, but once you start trading zucchinis with the store, you can’t expect to get the same price on sales to the store as you pay on purchases from the store.  The margin the store makes between the wholesale and retail price is what pays for the building, heating and cooling, labor, and other costs that are mostly fixed with respect to the amount you buy.

The same economics applies in electricity, only more so.  The retail price, especially in California, is covering a lot more than just the incremental cost of providing an extra kilowatt-hour to you.  In economic terms price is above the marginal cost of the incremental unit of energy, much further above than for goods you buy at the supermarket.  That price gap is paying for past losses from failed deregulation, costly nuclear power, expensive contracts with large scale renewables producers, and local distribution systems that carry power from the grid to your house, as well as metering consumption, billing and account collection.  As a result, when you consume less electricity, the cost the utility saves is much less than the revenue they lose.

Pricing above marginal cost means any decline in energy bought from the utility makes it harder to recover fixed costs, whether the quantity decline is from installing residential solar PV, improving energy efficiency, or just slow economic growth.  But buying less electricity for these reasons corresponds to growing your own zucchini (or just eating less zucchini).

Net metering of solar PV is equivalent to forcing the supermarket to take your zucchini on an even up trade for future zucchini.  The customer gets a one-for-one credit for electricity it puts into the grid against future consumption from the grid.  This expands the customer’s opportunity to save money, though in a way that reduces the utility’s net revenue.

This risk was largely ignored in the mid-1990s when net metering of residential solar PV was established.  It wasn’t that utilities or industry analysts failed to understand the simple math.  It’s that they didn’t think the exposure was very large, because solar PV was so expensive and the subsidies were smaller.  As recently as a decade ago, the cost of a residential system was still north of $10/watt, translating to at least $0.50/kWh.  Even with aggressive state subsidies and small federal subsidies, it was difficult to get the end-use consumer cost below $0.35/kWh.  The average retail price for the kWh replaced by a solar system was generally well below that, so very few consumers could really save money putting in solar.

But technology marches forward and PV panel costs have come way down.  Politics also marches (assign your own direction to it) and the effective subsidies for PV have increased substantially.  The war that is now erupting over tariff design is coming largely from utilities now taking distributed PV seriously.

A decade ago, utilities saw net metering as a small indirect subsidy to a nascent technology that was unlikely to ever be able to compete with even the retail price of electricity.  With plummeting prices of solar panels and some progress on installation costs, along with increased federal subsidies that now cover about half the cost, residential PV can now lower the retail energy bill of many California consumers, particularly customers of the large investor-owned utilities, which have the highest rates.

Two aspects of utility electricity tariffs are major contributors to the attractive economics of residential PV.  First, increasing-block pricing (“IBP”, higher price tiers as you consume more during the month) mean that for some heavy users, solar PV makes sense because it is replacing power on the highest tiers, where those consumers now pay $0.30-$0.40/kWh.  With lower costs and higher subsidies, PV has been able to beat those prices for at least a few years now.  That’s a major factor in the utilities pushing to flatten or eliminate IBP.

But the real panic in the industry has set in this year as the net-of-subsidy cost of PV has dropped below even the average retail price.  Even if increasing-block pricing were eliminated and the big IOUs sold all residential power for their average price of about $0.17/kWh, solar PV could beat that for many customers.  According to a recent report from Lawrence Berkeley Lab (and confirmed by other industry studies and media reports), the full cost of a typical residential system has fallen below $6/watt and may be below $5/watt by now, which corresponds to $0.25-$0.30/kWh.  The 30% federal tax credit and what’s left of the California Solar Initiative subsidies cuts that by about a third.  A less well-known tax effect – accelerated depreciation for leased systems – transfers another 15%-20% of the cost from the end-user to the federal government.  The net cost to the consumer can now be $0.15/kWh or less.

This is why we are hearing more often the phrase “death spiral” from the utilities.  If many customers act on the attractive economics of PV at home, the utility sells less energy and earns less above marginal cost to cover those costs of past sunk mistakes and ongoing fixed costs.  To make up the revenue, they would have to raise rates, which makes the economics of PV even better.

So, the utilities are now desperately pushing for tariff changes that a few years ago they saw as only a distant dream.  Not only do they want to eliminate increasing-block pricing, they want to further reduce the incremental energy price by implementing a fixed monthly charge on each customer, aimed at covering some of the costs of retail distribution, metering and billing.  Most economists support such changes as they move electricity pricing towards a more cost-based system.  In fact, the large publicly-owned utilities – Sacramento (SMUD) and Los Angeles (LADWP) – already have the tariff design the IOUs are now fighting for.  SMUD’s fixed monthly charge is $13, going up to $20 in a few years.

Not surprisingly, solar advocates love IBP and hate fixed monthly charges.  They argue that the proposed changes would hurt the poor – which is true if the changes aren’t accompanied by expanded discounts for low-income customers — though solar PV advocates don’t have a credible track record of protecting poor rate payers.  They also assert, with much less support, that solar PV adds so much extra value to the grid – by reducing line losses and the need for infrastructure upgrades – that solar should be  favored through the advantages that current tariff design gives them.

What makes the policy debate so difficult to resolve is that tariff design is a very indirect way to support residential PV.  In Germany, they’ve gone with feed-in tariffs for solar PV instead – a direct subsidy for every kWh of energy coming from your PV system.  Much simpler, and allows a reasoned debate on tariff design apart from solar PV policy.  But also makes it easier to see how much they are paying people to eat zucchini, and how that is driving up the bills of the people who prefer carrots (as made clear in a recent article in the center/left magazine Spiegel).

In the coming months, we are going to hear a lot of talk about tariff redesigns, solar PV penetration, and the utility business model.  Unless policy makers can separate rate design from residential PV policy, it’s not going to be easy to follow, or pretty to watch.



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.

45 thoughts on “Rate design wars are the sound of utilities taking residential PV seriously Leave a comment

  1. One of the implicit messages of the column is that the present system greatly but unevenly subsidizes solar PV. And as Bruce P says, much of the subsidy is almost invisible – it is not in the form of buybacks, but of avoided electricity purchases. One effect is that the magnitude of the subsidy varies depending on your particular location, tariff, etc. This is inefficient in principle i.e. we could get more solar for the same amount of total dollars, by using a more even subsidy such as the FIT.
    But as Severin is too polite to say, the current system is, perhaps inadvertently, well designed from a political economy standpoint. In order to install rooftop solar, you must a) own a house and b) have a good credit rating. In other words, it’s a subsidy for the well off, who tend to be well-informed voters. They will not favor a system that makes their subsidy obvious. So the main allies the utilities will have in this fight are dense city dwellers (no rooftops) and others who are stuck “eating carrots.” OTOH Jerry Brown is not afraid to attack tradition; perhaps the PUC will develop some fortitude.

  2. If the sole objective is to install as much rooftop PV as possible while ignoring the collateral damage, then feed-in tariffs are an excellent solution. The problem, of course, is that for all of its virtues, rooftop PV also leaves a fair amount of collateral damage in its wake. As the Germans and their neighbors have discovered, it plays havoc with grid operations and it exacerbates social inequity by transferring wealth to those who can afford PV from those who cannot. Spain has also built a lot of rooftop PV, and as I recall, the huge cost of Spain’s FIT subsidies were still sitting on the balance sheets of the country’s major utilities with limited prospects for recovery, which in effect means those firms are insolvent. Also keep in mind that allegedly “clean” PV uses materials that are at least as toxic as long-lived as radioactive spent fuel yet PV owners are not required to set aside a penny to deal with safely disposing of PV panels at the end of their lives. Finally, even if Tam is correct about the $3 billion annual cost for California associated with adoption of feed-in tariffs, that’s still a fairly significant sum in a $40 billion dollar retail electricity market.

    The California legislature, at least two governors and the CPUC have put themselves in a box over renewable energy and as Severin points out, there is no painless way for them to escape other than to hope they’re termed out before voters wise up

    Although it isn’t a perfect solution and it won’t help politicians escape the web they’ve woven around themselves, I would do away with net energy metering and prohibit feed-in tariffs in order to avoid the cross-subsidies and economic distortions that come with them. Instead, I would require that every connected customer pay a volumetric fee for the use of the wires in either direction. Customers would also pay the time-dependent (hourly or more frequent) wholesale energy price for energy they take from the grid, and they would be paid the wholesale price for energy they deliver to the grid. Customers and their energy providers would also be free to enter into alternative arrangements outside the control of the utility regulator to hedge the cost of energy they consume and any revenues for surplus production. No doubt deriving rates for the use of the wires raises all sorts of other issues, but they pale in comparison with the havoc brought on by feed-in tariffs and net energy metering.

    • Jack, a few clarifications: 1) the $3 billion figure I cite is the E3 report’s estimate for the cost of net metering, not for FITs. 2) The cost-effective FIT that I advocate would, by definition, have zero additional cost b/c this is what I mean by cost-effective. We already have avoided cost policy structures for PURPA and for the RPS program (the “Market Price Referent” which is a species of avoided cost). 3) Net-metering and FITs are mutually exclusive in CA, so there is no “cross-subsidy.” You choose one or the other. 4) Germany has had minimal issues integrating high levels of PV. They have achieved over 40% of their power coming from PV a number of times this year and there have been no grid problems. And they’re going full steam ahead in installing more PV, to the tune of 7 GW in each of the last three years and probably about half of that in 2013. Germany’s FIT rates are now below the retail cost of power, so they have past the point of grid parity. This means that it makes no sense for a business to not install solar power b/c if they don’t they’re going to pay more for power from the grid than they would through consuming solar power on-site and sending any excess to the grid. Germany has paved the way for other countries to rapidly decarbonize their electric grids in a cost-effective manner.

  3. Imagine you stop shopping at a store for zucchini because you grow your own, and you give some to your neighbors. Then imagine the store sends you a bill each month for every day that goes by, because they have a store available to you that you no longer use, and that the store also charges your neighbors for the zucchini you gave to them.

    • Now imagine the store closes, but there isn’t enough zucchini to feed your neighbors. Good think you have a garden and a gun to protect it since there is no longer a store to work at and your neighbors are starving. But they won’t last much longer, so as long as you got yours all is good. Right?

      Over 35% of all residential customers rent and almost all commercial customers do so. On-site power isn’t an option for them. Even with “community solar” they will need the utility to get power to them. Killing the utility industry isn’t the solution, yet that is the likely outcome if these policies keep shifting costs to non-participants. If this policy continues, municipalization is inevitable to close the loop between ratepayer and taxpayer funding.

  4. Great article. The underlying problem is that the electricity bill reductions achieved by solar PV customers exceed the costs that utilities avoid as a result of that behind the meter generation. (The bulk of those avoided costs are wholesale power purchase costs andd/or fuel costs).

    In order to recover the remainder, utilities have to charge other customers higher prices.

    However, it should be borne in mind that, at least in California and Arizona, the amount of energy solar PV customers consume for themselves usually significantly exceeds the amount of genetration they sell back to the grid. That is especially true for residential customers.

    Therefore most of the cross-subsidy solar PV customers receive from other customers that are charged higher prices to recover costs that are not avoided by “behind the meter” solar PV generation is due to the electricity that solar PV customers generate to meet their own load. Energy sold back to the grid by solar POV customers accounts for only as small portion of that cross subsidy.

  5. Great discussion, and Severin is right about what has created the return of the ‘death spiral’ talk of the late 1980’s. But don’t forget that the current system has created massive amounts of externalized costs and should not be used as a framework. The future has to be very different from the past. To use the analogy, we need a new system with everybody producing and eating as much zucchini as possible. The real question is what can we do to get there even faster? Our children are depending on us to get that done.

  6. Incenting solar (or any above marginal cost resource, including small nuclear units) for public policy reasons isn’t unreasonable – to a point. However, it conflates utility customers paying the explicit and implicit subsidies with “citizens” and the taxpayers who would otherwise have to pay the additional costs (and to some extent do). Like all public policies, at some point the policy should be evaluated against its goals and terminated or modified.
    It seems clear the PV industry has matured to the point where it can provide power competitively with retail rates. It may be less clear that it can thrive without continued public support, although the exact form of that support is appropriate to debate at this time. To the extent it simply displaces customer purchases from the utility it is assumed to be no different than conservation; however, I would argue self-generation, especially exporting power back to the grid is a difference in kind not of degree. From a rate design perspective retail customers selling back to the utility aren’t the “same” as other members of that customer class. Therefore equity reasons, they should be assigned to a different customer class and allocated the appropriate costs associated with serving them to hold “nonparticipants” harmless.
    A class division on the basis of generation may not be the end. As more retail customers engage in ancillary service markets using DR and/or storage likely present similar equity concerns, at least to the extent the dispatch of DR and storage is NOT controlled by the retail utility, but by the customer or the RTO. Both situations have in common their disconnection from the operation of the retail utility, which is forced to adapt to; and adapt outside the range of load uncertainty assumed for “passive” customers. This may recommend assigning some DR and storage customers to the new “active customer” class.

  7. Severin, interesting piece and I like your suggestion that feed-in tariff policies are far simpler and straightforward than rate design and net metering. I’ve made the same point for years in various policy circles and the problem is that there is such an entrenched support system for net metering that it’s hard for people to consider these issues objectively.

    When we look to the scale of what Germany has achieved with a simple FIT vs. California’s net-metering approach (35 GW in Germany vs. about 1 GW in CA) it should be obvious that the FIT wins hands-down – particularly now that Germany has almost singlehandedly brought the cost of solar PV to below grid parity. And that’s my second point: you suggest that Germany’s FIT is a subsidy, but it’s not anymore. It certainly was for a few years but the payments now offered under the subsidy are LESS than what retail power costs now in Germany (a recent report from Germany’s Ministry of Trade found that grid parity was reached in 2011 in Germany). And this has happened entirely b/c of Germany’s FIT program and China’s response to the FIT program (dramatic increases in production of solar panels).

    Last, you mention but don’t discuss the value that PV provides to the grid. This is far from a negligible value, with one study by E3 for the CPUC finding that in some locations this “locational benefit” can be as much as 6 c/kWh, which is huge. I also note that the recent report to the CPUC, also from E3, on the cost of net metering for ratepayers, found that the net cost by 2020 of achieving the current 5% system cap would be only about $3 billion per year, which doesn’t include the grid benefits, and this is quite a small figure considering how much we all spend on energy in CA.

  8. First, I have solar PV on my roof (SMUD), and have had it for over 12 years (and I already pay ~$12.50/month in fixed costs).

    I wish you had addressed time-of-use costs and value. During the day, my system is providing electrons to SMUD. Then, during the evening, I am pulling electrons off the grid. The grid is a “virtual battery” for me as I think if it. SMUD gets the electrons when the cost of electricity is high and I use the electrons when the cost of electricity is low. During the summer I am a net provider of the electrons, and during the winter I am a net consumer.

    I understand that any retailer cannot buy their inventory (including the electrons) at retail and still pay their costs, and get any return. However, as a SMUD customer, I have the same incentive to conserve electricity whether or not I have the PV system, because the electrons have the same value regardless of the direction of electron flow.

    I would like it if you could include time-of-use valuation in your assessments!

    — Gerry Bemis

    • Gerry,

      In my comments just before yours, I say that people like you, who provide electricity to the grid when the price is high and take electricity from the grid when the price is low act like a battery for the grid, not the grid acting as a battery for you.


      • I think all utilities would gladly pay the value of the energy. Your point is valid that daytime prices are higher than night time. But, the value of the energy on peak is typically 3-5 cents per kWh; Not the 12-18 cents in many rates. Some tiered rates reach as high as 45 cents. (Energy at night ranges closer to 1 to 4 cents.)

        Much of the fixed costs of your energy supplier are due to maintaining the capacity available to meet the demand on the peak hours. At time of peak, the output of the solar panels is greatly diminished. For referece, look in the CAISO projected load curves.

        I contend that the best way to charge the customer is a demand charge and an energy charge. Such a paradigm would give economic incentives to control the simultaneous use of electricity. The problem is, that I believe few people would turn off their AC and refrigerator on the hottest of days. That is a legitimate choice, but it should have appropriate economic consequences.

      • Jim,
        Apparently your comment advocating demand charges got stuck in my subconscious because three months later I was asked to speak at the Deloitte Energy conference in May 2014 and offered to speak on the topic of rate reform, including a residential demand charge. This lead to additional presentations on the topic and two published papers, “Curing the Death Spiral,” with Lori Cifuentes (Tampa Electric Company), Public Utilities Fortnightly, 2014 August ( and “Demand a Better Utility Charge During Era of Renewables: Getting Renewable Incentives Correct With Residential Demand Charges,” Dialogue, United States Association for Energy Economics, 2015 January (

        A Google search brought me back to this blog a few minutes ago and my belated realization that I needed to thank you for your inspiration in regard to residential demand charges. Send me your reaction to the two papers.


  9. Severin,

    Nicely written article. But you left out part of the uniquely California solution, mandatory storage. The store that won’t buy the zucchini refuses to provide storage at no cost to the consumer. Or is it the other way around, the customer is providing storage to the store. A lot depends upon the value of the zucchini/electricity when it is being delivered and returned.

    We put electricity into storage when the electricity is low priced. We take electricity out when the electricity is high priced. If a customer delivers surplus electricity during high priced periods and then takes the electricity back during low priced period, it can be argued that the system is storing electricity with the customer, or it is the equivalent of storage by the customer.

    With Advanced Metering Infrastructure, we can price the ins and outs on fifteen minute periods (or even shorter), reflecting the high and low marginal costs associated with the operation of the California ISO. This cashes out some of the storage aspects of the zucchini, but still leaves the huge regulatory costs you mentioned.

    AMI and the smart grid might allow time differentiation of the use of the wires, as I discuss in “Dynamic ‘Distribution’ Grid Pricing,” a draft of which is on my web page, The inspiration for this paper was a speaker who said that some Hawaiian distribution feeders have three times the distributed generation as they have load. The voltage problems are unusual and need an unusual price structure. Though “Dynamic ‘Distribution’ Grid Pricing” was inspired by the October presentation I attended, it is in some respects a reprise of “Microgrids And Financial Affairs – Creating A Value-Based Real-Time Price For Electricity,” Cogeneration and On-Site Power Production, September, 2007.

  10. To me, this demonstrates how totally out of control the subsidies for solar have become. Net metering would still be a non-issue if not for the massive amount of subsidy lavished on Solar. Absent the subsidy, solar would likely be not cost effective and an inconsequential contributor to the electric transmission system.

    The sad thing is that consumers without solar foot the bill for consumers with solar. Solar consumers tend to be affluent and least in need of subsidies. So these subsidies are basically a transfer of wealth from less affluent consumers to more affluent consumers. The authors rejoinder is that we will give subsidies to least affluent people since they are footing the bill for the most affluent people. So who foots the bill for those subsidies? Since the most affluent are subsidized and the poor are subsidized, that leaves only those in between the two. I ask why they ultimately get stuck with the bill for both the rich and the poor?

    • Jardinero, the subsidies for solar amount to a tiny fraction of your utility bill. The Public Goods Charge, which used to fund solar subsidies as well as energy efficiency programs and some other items, amount to 1-3% of the average utility bill. And now that solar subsidies are going away, per the planned reduction in subsidies over time, this is a moot point. Solar also receives a 30% federal tax credit and accelerated depreciation on the wholesale side or if a business installs solar, and one could argue that these are subsidies. However, one could argue equally coherently that these are simply allowing the owner to keep more of their revenue from the project instead of paying that in taxes. Last, most energy projects, including nuclear, receive very similar or even better tax treatment under federal rules. It is certainly the case that solar power has received substantial subsidies in the past, but, again, these subsidies are a tiny fraction of what ratepayers pay, and they’re going away b/c they’ve accomplished their mission: bringing costs down to the point where subsidies aren’t required anymore. And this is an amazing success story that is transforming our planet for the better as we speak. In two decades, most Western countries will obtain a majority of their power from solar, wind, geothermal, biomass, small hydro, and energy storage. The rest will probably from natural gas and nuclear.

%d bloggers like this: