Without major rate reform progress will be slow and incomplete.

Not everyone has yet agreed that rooftop solar systems in California’s investor-owned utility territories shift significant costs onto other customers. Nonetheless, some legislators and regulators in the state seem ready to face this reality as part of trying to address the affordability crisis in electricity bills. There is now an active discussion in Sacramento about how to reduce the cost shift and a lot of ideas have surfaced. So, I thought I would discuss the ones I have heard about and invite readers to suggest others.

Spoiler alert: It’s really hard to fix the rooftop solar cost shift without fixing rates.

The NEM phaseout. When systems were installed under NEM 1.0 (which was in effect until around the end of 2016) and NEM 2.0 (approximately 2017 to April 14, 2023), the regulations said that customers would be covered by those compensation policies for 20 years. Some legislators have suggested shortening that to 10 years, after which these customers would be covered under the Net Billing Tariff (NBT, also known as “NEM 3.0”), which went into effect for new systems installed after April 14, 2023. 

NBT is net metering over an hour rather than a year. Most estimates suggest that under NBT a typical residential solar customer without battery storage would still receive credit based on the retail rate (40-60 cents per kWh these days) for about half of their rooftop generation. For the other half, the credit would reflect the CPUC calculation of avoided cost for exports to the grid (about 6 cents), which by the definition of avoided cost does not create a cost shift.

The rooftop solar industry and many of the customers covered by these potential regulatory changes are  outraged, saying that the change would break a contract. I’ll leave it to the lawyers to debate whether the regulatory language is a contract. Still, the same folks also argue that such a change would be unfair to early solar adopters. I’m not sure the fairness argument really works. 

A customer who put in rooftop solar in 2015 on average paid $5,460 per kilowatt of capacity before the 30% tax credit. If they didn’t oversize their system (and few people did back then) then they’ve been able to earn full retail credit for all of the kWh production of their systems. Under reasonable assumptions about output and rates, the figure below shows the aggregate net present value by year of the savings from installing solar, net of the upfront cost.

_{Source: Author calculations assuming that (1) 4 kW system installed at the beginning of 2015 at a cost of 4*5460=$21,840 before 30% tax credit, which lowers the cost to $15,288. (2) System capacity factor starts at 20% and declines by 0.2 percentage points per year. (3) Retail rate avoided is non-CARE average retail price up to 2024, after which retail price is assumed to be constant in real terms. (4) All monetary values in 2015 dollars. Assumed interest rates are real, above the rate of inflation}.

The bottom line is that the average customer who installed a system in 2015 hit the breakeven point around 2023 to 2027 and benefits continue to accrue for years beyond. Households who adopted solar after 2015 have seen their investments pay off faster. They paid somewhat less for their systems and will avoid much higher retail prices during the first 10 years of their systems. The slight decrease in compensation under NEM 2.0 – having to pay a non-bypassable charge of 2-3 cents/kWh (for systems installed in 2017 and later) – is more than offset by the higher rates avoided.

So, I think this is a fair proposal. But I also think it wouldn’t do much to reduce the cost shift. Over the next 8 years, the proposal would gradually convert all NEM 1.0 and NEM 2.0 customers to NBT, but NBT still retains about half of the cost shift. 

More importantly,   with a battery and an AI-enhanced algorithm, a solar household under NBT can effectively “self-NEM” and get full retail credit for much more than half of their rooftop generation, thus paying very little more than a customer under NEM 2.0. The vertical distance between the solid and dashed lines in the figure suggests how much value they could get from self-NEMing. 

(Source)

To be clear, rooftop solar customers are not breaking any rules or doing anything unethical. It’s not their job to figure out equitable rooftop solar policy. That’s the job of regulators and legislators.

Eliminating the Climate Credit for solar households: Under the legislation that created the California cap and trade market for greenhouse gases, households receive a semi-annual bill credit based on the revenues the state gets from the auction of cap and trade allowances. These days, a typical household is getting $80-$100 per year off their electricity bill (which averages over $2400 per year for those without solar who are not on the low-income program, CARE). One proposal is to eliminate this credit for households with rooftop solar.

This is one way to claw back some of the excessive subsidy from net energy metering, but I foresee some objections.  First, it takes back the same amount from all solar households regardless of the size of the solar subsidy that the household is receiving. If the price of allowances rises – as seems likely if the program is extended beyond 2030 – it is even possible that the annual climate credit could be worth more than the cost shift from a household with a small solar system.

Second, there isn’t an obvious nexus between the climate credit and the cost shift. Nearly all households with solar do still pay for some electricity from the grid, just less than other households, and much less than they would have otherwise purchased. So defenders of rooftop solar will argue that the original intent of the climate credit – to rebate some of the additional electricity cost due to cap and trade – also applies to them.

Converting from NEM 1.0 or 2.0 to NBT if a house changes ownership:I’m not sure what the justification for this would be. Perhaps it’s the idea that the owner who made the investment is no longer benefiting once they sell the house. Such reasoning, however, ignores the fact that selling the house with NEM 1.0 or 2.0 eligibility will make it worth more than if it is under NBT.

My main concern is not with the impact on electricity, but on housing markets. California already puts a lot of frictions into house sales, primarily through the favorable property tax treatment that longtime residents unfairly receive under Proposition 13 (myself included). That encourages owners to stay put, even if their house or location is no longer a good fit for them. The last thing the state needs is more policies that discourage moving.

Changing NEM 1.0 and 2.0 subsidies to be based on retail price at the time the system was installed (adjusted for inflation): The idea here is that households that installed solar 5-10 years ago did so without expectation of the astronomical rate increases over the last five years caused by grid hardening costs, wildfire liabilities, the rooftop solar cost shift, and other factors. Those increases have raised the effective subsidy to rooftop solar, an unexpected windfall for solar owners. So, why not calibrate the subsidy to an estimate of the expected future retail prices at the time the system was installed?

This has some appeal on equity grounds. But, practically it has some issues. It would only apply to exports to the grid, because it would not make sense to base the solar household’s cost of purchases from the grid on the earlier retail price. Compared to NBT, this approach would reduce the customer’s incentive to use a battery primarily to “self-NEM”, because the customer would receive a higher compensation for exports than under NBT. But it would also reduce the cost shift by even less than NBT.

Fix Retail Rates: For years, I and others have argued that prices should reflect social marginal cost (SMC), but that economic logic has gotten little political traction.

(Source)

So here is another approach: In 1994 – before California started electricity restructuring, before the renewable portfolio standard or significant investments in renewables, and before climate change completely transformed wildfire risk – the average residential rates of PG&E, SCE and SDG&E exceeded the national average by 45%, 46%, and 30% respectively. In 2023 (the most recent year for which EIA has published the data), the differences were 122%, 111%, and 196%, respectively. 

The great majority of that relative price increase is due to the fixed costs of our fight to adapt to climate change, to pay for damage from past fires it has caused, and to support policies that mitigate future climate change. If those additional costs were paid by the state budget, so that 1994 percentage differentials remained, PG&E’s average 2023 residential rates would have been $0.22 instead of $0.34, SCE’s would have been $0.22 instead of $0.32, and SDG&E’s would have been $0.20 instead of $0.45. The normal fixed costs – poles and wires, billing and customer service, etc. – would still be in the rates.  And the 2023 cost shift would have been reduced by roughly 45% to 70%. 

Reality Check

Of course, lowering rates by moving costs to the state budget doesn’t make those costs disappear. We also need to continue to reduce the costs of responding to climate change, as well as all of the normal costs of utility operations. And paying those costs from the state budget means they fall on taxpayers, rather than ratepayers. But it also means that they don’t fall disproportionately on low income households, and they don’t discourage electrification as our current rates do. Lowering rates is politically difficult, but it would substantially lower the solar cost shift, unlike the other policies I’ve heard proposed.

—

The Energy Institute blog will be on vacation next week for Memorial Day. It will return on Monday, June 2.

I am now posting suggested energy readings (and some political views) most weekdays on Bluesky @severinborenstein.bsky.social

Follow us on Bluesky and LinkedIn, and subscribe to our email list to keep up with future content and announcements. 

Suggested citation: Borenstein, Severin. “How To Fix the Solar Cost Shift” Energy Institute Blog, UC Berkeley, May 19, 2025, https://energyathaas.wordpress.com/2025/05/19/how-to-fix-the-solar-cost-shift/

—

By default comments are displayed as anonymous, but if you are comfortable doing so, we encourage you to sign your comments.