The short answer: cleaner power did not automatically mean cheaper or self-sufficient power
California’s transition produced a striking combination in 2024. The average retail electricity price reached 27.04 cents/kWh, more than twice the U.S. average of 12.94 cents/kWh, while the carbon intensity of in-state power generation was 407 lb/MWh, far below the national 785 lb/MWh. [1][10] Those two numbers rule out two overly simple stories. Cleaner generation did not guarantee low retail rates, and high retail rates cannot be attributed to carbon-free resources alone.
The supply picture is equally mixed. CEC counted 278.3 TWh in the 2024 total system mix, including 216.2 TWh of in-state generation and 62.2 TWh of net imports. Clean resources supplied 62% of that mix, and behind-the-meter solar capacity exceeded 17.4 GW. [2] California therefore became cleaner without becoming electrically isolated from the rest of the West. Import dependence must be evaluated alongside hydrology, storage, demand and the hourly shape of net load, not just solar nameplate capacity.
This report treats four outcomes separately: decarbonization, retail price, interstate imports and household affordability. They interact, but they use different accounting boundaries and respond to different mechanisms. Keeping them separate is the only way to identify where the transition has delivered clear gains and where the remaining problem is primarily one of system operation or cost allocation.
Accounting boundaries: generation, retail sales and total-system supply are not interchangeable
EIA reports state generation, retail sales, prices and power-sector emissions. In 2024 California produced about 214.2 TWh of net generation but recorded roughly 245.7 TWh of retail sales. [1] The gap is evidence that in-state generation alone does not cover state electricity use, but it is not a plug-in estimate of imports because direct use, losses and dataset boundaries differ.
CEC’s Total System Electric Generation series is designed to combine in-state generation with imports serving California load, while CARB’s greenhouse-gas inventory explicitly assigns emissions to imported electricity. [2][3] CAISO data are different again: they describe the ISO balancing area at five-minute resolution and are suited to operational questions such as midday surplus, evening ramps and battery dispatch. The datasets can be used together only if their geography, time scale and denominator remain explicit.
Accordingly, this report uses EIA for state price and generation-carbon comparisons, CEC for annual system supply and imports, CARB to define the consumption-side emissions boundary, and CAISO for hourly operating mechanisms. It does not collapse those measures into a single index.
The 2024 baseline: high retail prices coexisted with relatively low generation emissions
The California-U.S. comparison is the simplest demonstration that price and generation emissions are separate outcomes. California’s average retail price exceeded the national average by 14.10 cents/kWh, yet its power-sector carbon intensity was only about 52% of the U.S. level. [1][10] The descriptive result does not prove what caused the price gap, but it shows that generation carbon intensity by itself is not a sufficient explanation.
Western comparators strengthen the point. Washington combined 10.13 cents/kWh with 249 lb/MWh and Oregon 11.11 cents/kWh with 352 lb/MWh. Arizona and Nevada were also much cheaper than California, although their generation was more carbon intensive. [11][12][13][14] Resource endowments, legacy hydro, ownership structures, wildfire exposure, transmission investment and regulatory design all differ. The comparison is therefore diagnostic rather than a ranking.
| Region | Average retail price (¢/kWh) | Generation CO₂ intensity (lb/MWh) | What the pairing shows |
|---|---|---|---|
| California | 27.04 | 407 | High price, relatively low carbon intensity |
| United States | 12.94 | 785 | National benchmark |
| Washington | 10.13 | 249 | Low price and very low carbon intensity |
| Oregon | 11.11 | 352 | Low price and low carbon intensity |
| Arizona | 12.74 | 634 | Lower price, higher carbon intensity |
| Nevada | 11.47 | 631 | Solar-rich system with a different cost structure |
How large did solar and clean power become?
CEC reports that clean resources supplied 62% of the 2024 total system power mix. On a retail-sales basis, RPS-certified renewables supplied 45.2%, while the broader clean category including large hydro and nuclear reached 67%. [2][19] These percentages are not supposed to match: the denominator and resource definitions differ. Their common signal is that California’s supply mix has moved decisively toward low-carbon resources.
Behind-the-meter solar is now large enough to change how utility demand is interpreted. CEC reports more than 17.4 GW of BTM solar in 2024 and estimates that it displaced roughly 10% of energy that otherwise would have been delivered by local utilities. [2] A decline in utility-delivered load therefore cannot be read as an equal decline in electricity services consumed by homes and businesses. Distributed generation increasingly sits outside the traditional utility-sales meter.
Policy created much of this trajectory. The RPS drove renewable procurement over two decades, and SB 100 requires renewable and zero-carbon resources to supply 100% of retail sales by 2045. [24] The planning challenge has now shifted from simply adding clean megawatts to integrating generation with storage, transmission, reliability and changing demand.
Why imports still matter even with a much larger solar fleet
CEC counted 216.2 TWh of in-state generation and 62.2 TWh of net imports in the 2024 total system mix. Imports therefore represented roughly 22% of system supply. [2] That does not make imported electricity inherently fossil-based: Western imports can include hydro, wind, nuclear and gas. It does mean that California’s power balance still depends materially on resources outside its borders.
The emissions boundary matters for the same reason. EIA’s 407 lb/MWh figure describes in-state generation, whereas CARB includes emissions associated with imported electricity in the state inventory. [1][3] “California’s generation is relatively low-carbon” is well supported. “All electricity consumed in California has the same low carbon intensity” is a different claim and requires information on imported power.
Imports fell after 2022, but solar is not the only explanation
CEC data show net imports at about 83.6 TWh in 2021 and 84.0 TWh in 2022, falling to 65.5 TWh in 2023 and 62.2 TWh in 2024. [7][8][9][2] The decline from 2022 to 2024 was roughly 21.8 TWh. Solar and storage growth likely contributed to lower net import needs in some hours, but the annual change cannot be assigned to solar alone.
Hydrology is the clearest confounder. California hydro output weakened sharply during the 2021 drought, while wetter conditions in 2023 lifted hydro generation by roughly 80% from 2022. [7][9] Regional hydrology also affects what is available to import. CEC reported that drought in British Columbia during 2023-24 reduced imports from BC Hydro and BPA by about 70% relative to 2022 levels. Annual imports therefore reflect California generation, regional hydro, demand, outages and market prices at the same time.
Hourly operations explain why solar growth and higher imports can occur at the same time
Power systems balance every few minutes, not once a year. CAISO describes a familiar pattern: abundant midday solar depresses net load, then solar output falls rapidly in the late afternoon while demand remains high. Gas units, batteries and imports must ramp to fill that interval. [15] The operational question is therefore not only how much solar energy is produced, but whether flexible resources are available when solar output is falling.
The January-May 2026 comparison provides an unusually clear illustration. EIA reported that CAISO utility-scale solar generation was 21% above the same 2024 period and gas generation was 60% lower. Demand was 7% higher, CAISO net generation was 19% lower, and imports approximately doubled. [16] Those facts do not mean solar caused imports to rise. They show that demand, hydro, outages and regional supply can dominate the import balance in particular operating periods even while solar generation expands rapidly.
This is why annual import dependence should be supplemented with an hourly metric. Midday exports or curtailment can coexist with evening import needs on the same day. Storage changes the connection between those hours by moving energy rather than creating additional annual generation.
Gas and batteries are primarily about timing, not an ideological opposite to solar
CAISO continues to rely on gas and imports because their dispatchable timing differs from solar output. Solar is abundant at midday; the evening system needs energy after the sun has set. Batteries increasingly bridge the interval, but duration, state of charge and transmission constraints determine how much firm evening capacity they can replace. [15] A system can therefore have a high annual solar share and still need flexible gas or imports during specific hours.
Planning documents increasingly treat distributed storage together with BTM solar rather than as a separate niche. CEC forecasts BTM PV and storage through 2050 and includes rates and self-generation in its demand models. [20][23] That shift matters: the next phase of California solar policy is about when electricity is delivered and what grid service it provides, not only how many gigawatts are installed.
Retail rates include much more than the cost of generated electricity
California retail rates recover generation procurement, transmission and distribution investment, wildfire mitigation and restoration, insurance and liability costs, public-purpose programs and other authorized revenue requirements. CPUC rate and utility-cost reports explicitly link rate changes to general rate cases, generation adjustments and wildfire-related spending. [4][5] A low wholesale solar contract can therefore coexist with a higher final retail rate if other revenue requirements rise faster.
Bill credits add another distinction between underlying cost and what a customer pays in a particular period. CPUC reports about $1.89 billion in electric cap-and-trade bill credits in 2024. [4] Credits reduce customer bills, but they do not erase the utility revenue requirement that is being recovered through the rate structure. For affordability analysis, both layers matter.
Wildfire spending and legacy rooftop-solar cost allocation are material rate issues
CPUC evidence shows that wildfire-related costs have become a large part of major investor-owned utility revenue requirements. One comparison put the share at roughly 27% for PG&E and 17% each for SCE and SDG&E. [5][6] The category includes mitigation, hardening, rebuilding, insurance and liability-related recovery. Climate risk therefore reaches electricity bills through infrastructure and finance as well as through generation conditions.
Legacy rooftop-solar compensation is another distinct cost-allocation issue. CPUC has documented a large annual cost shift associated with older net-metering structures, which is one reason the state replaced NEM2 for new applications with the Net Billing Tariff. [5][17] This does not imply that rooftop solar produced no capacity, emissions or local-generation benefits. It means the value of exported electricity and the recovery of fixed grid costs have to be designed explicitly.
| Cost or policy channel | What official evidence supports | What it does not establish by itself |
|---|---|---|
| Wildfire and resilience | Double-digit shares of IOU revenue requirements | That clean generation caused wildfire costs |
| Generation procurement | Fuel, contract and market costs affect rates | That lower wholesale costs translate one-for-one to retail rates |
| Legacy NEM allocation | A cost shift to nonparticipants is documented | That all rooftop-solar benefits are smaller than costs |
| Transmission and distribution | Grid investment is recovered through rates | That every investment has the same customer value |
| Cap-and-trade credits | Credits can reduce bills in a given year | That the underlying revenue requirement disappeared |
Household affordability is not the same thing as the statewide average price
The statewide average is useful for comparison but poor as a household-burden measure. Customers in hot inland climate zones can consume far more electricity than coastal customers, and the same dollar bill imposes very different stress at different income levels. CPUC therefore uses affordability-ratio tools that relate essential utility costs to household resources rather than relying on cents/kWh alone. [22]
Program eligibility can materially change the bill. CARE provides roughly 30-35% electric discounts for eligible customers and FERA provides an 18% discount; qualifying customers also receive lower fixed charges under the new structure. [18] A household’s exposure therefore depends on utility territory, tariff, consumption, income eligibility, fixed charges and whether it owns solar or storage.
CEC’s low-income bill analysis reaches the same methodological conclusion from another direction: electrification and decarbonization measures can have different bill effects across climate regions, building types and rate conditions. [21] Any claim about “the California household” should therefore be treated with caution unless the tariff and household profile are specified.
Western-state comparisons show that low carbon does not mechanically imply high price
Washington and Oregon provide useful counterexamples to the idea that low-carbon electricity is inherently expensive. Washington combined just 249 lb/MWh with a 10.13 cent/kWh average retail price, and Oregon combined 352 lb/MWh with 11.11 cents/kWh. [11][12] Their large hydro fleets and different utility structures are not directly reproducible in California, but the comparison demonstrates that carbon intensity is not a sufficient explanation for California’s price level.
Arizona and Nevada make the opposite point: both were much cheaper than California while generating electricity at substantially higher carbon intensity. [13][14] Price, carbon and resource mix are linked through many institutions and assets rather than a single technology. State comparisons should therefore be used to test simplistic causal stories, not to declare a universal best model.
Policy has shifted from maximizing solar exports toward valuing time and integration
Early policy emphasized renewable procurement and rooftop adoption. As midday solar became abundant, the marginal value of an exported kilowatt-hour became more dependent on timing. The Net Billing Tariff, effective for new interconnection applications from April 2023, aligns export compensation more closely with grid value and strengthens incentives to pair solar with batteries. [17] The policy change reflects a new system need rather than a rejection of the earlier build-out.
The broader SB 100 process is moving in the same direction. The 2045 clean-retail target is evaluated together with reliability, affordability, transmission, technology and project-development uncertainty. [24] California’s next transition problem is therefore integration: matching clean production to demand in time and location while controlling the non-generation costs embedded in retail rates.
Key findings: a four-axis synthesis of the transition
Separating the evidence into four axes produces a clearer assessment. Decarbonization shows strong progress. Retail price remains a high-pressure outcome. Annual import dependence has declined from 2021-22 levels but remains material and can rise in specific operating periods. Household burden is highly uneven because income, tariff, climate and discount eligibility matter. These outcomes did not improve at the same speed.
The framework is not a scorecard and does not rank California against other states. Its purpose is diagnostic. A problem on the import axis may require storage, transmission or demand flexibility; a problem on the affordability axis may require cost control and tariff redesign even if generation is already cleaner. Keeping the axes distinct makes policy responsibility easier to identify.
Counterevidence and alternative explanations
The first counterpoint is that the correlation between more solar and higher California retail rates is not enough to establish solar as the dominant cause of the rate increase. CPUC identifies wildfire, transmission and distribution investment, liability and insurance, generation adjustments and legacy NEM cost allocation as separate drivers. Solar policy is part of the cost story, but official evidence does not support treating it as the entire story.
The second counterpoint is operational. Early-2026 CAISO conditions showed solar up sharply while imports also rose. [16] That observation does not disprove a long-run import-reducing effect from solar and storage. It shows that the effect depends on demand, hydro, outages, battery duration and the specific hours being examined. The same caution applies in the other direction: the 2021-24 import decline should not be credited wholly to solar.
A third alternative explanation is resource endowment. Hydro-rich states can achieve low prices and low carbon intensity using legacy assets and hydrologic resources that California cannot simply copy. Cross-state evidence is strongest as a falsification test for one-factor explanations, not as a template for identical policy.
Forward scenarios: what would lower imports or improve affordability?
Import dependence should fall further if batteries move more midday solar into the evening, transmission adds access to diverse resources, efficiency and demand response reduce peak requirements, and hydrology is supportive. It can remain high or rise in some hours if data centers, transport and building electrification increase load faster than flexible supply or if Western hydro weakens. CEC’s 2025-2045 demand forecast explicitly models rates, BTM PV, electrification, efficiency and data-center load together. [20]
Retail affordability follows a different scenario tree. Lower fuel exposure can help over time, but customer rates may remain under pressure if wildfire and network revenue requirements continue to rise. Conversely, better cost control, efficient grid investment, bill credits and targeted low-income programs can reduce household burden even before the statewide average rate falls materially.
The most favorable transition therefore combines three things rather than one: abundant clean energy, enough flexibility to deliver it when demand is high, and a cost-recovery design that does not let non-generation revenue requirements overwhelm the energy-cost savings.
Limitations: what this evidence cannot isolate
This report does not estimate a single causal coefficient for solar and retail rates. Doing that credibly would require utility-level cost ledgers, procurement contracts, grid investment, wildfire spending, load growth and policy costs aligned over the same period and accounting rules. The public sources used here establish important magnitudes and mechanisms, but they are not one econometric panel.
The report also does not reconstruct the fuel mix and emissions intensity of every imported megawatt-hour. CARB provides the appropriate consumption-side boundary, but the annual 62.2 TWh import total is not enough to identify what generation was displaced in every hour. In-state generation carbon intensity is therefore kept separate from consumption-related electricity emissions.
Finally, the 2026 CAISO comparison covers January through May rather than a full year. It is used as current operational evidence, not as a replacement for the 2021-24 annual trend. The distinction matters because hydro, demand and outage patterns can change sharply over the remaining months.
Practical implications
Performance reporting should use separate dashboards for generation carbon intensity, retail price, import timing and household burden. A single clean-energy percentage can hide affordability problems, while a high retail rate can obscure genuine emissions progress.
Rate reporting should separate generation procurement from wildfire, transmission and distribution, legacy NEM allocation and other revenue requirements in a common format. That would let customers and policymakers compare the cost of clean-energy procurement with the cost of operating and hardening the wider system.
Solar policy should measure storage pairing, evening discharge, curtailment avoided and imports displaced by hour rather than celebrating installed capacity alone. Those metrics are better aligned with the system value that the Net Billing Tariff is intended to encourage.
Affordability policy should track essential-service burden after CARE/FERA discounts and fixed charges, not just average cents/kWh. The distribution of costs across income groups is a separate policy outcome and should be managed as such.
Conclusion
Three findings are firm. California’s in-state electricity generation is much less carbon intensive than the U.S. average; solar and storage have materially changed the supply system; and retail prices remain exceptionally high while imports remain significant. These facts coexist because they describe different parts of the electricity system.
Two outcomes are conditional. Import dependence varies with hydrology, Western supply, evening net load, battery duration and demand growth. Household affordability varies with usage, utility territory, income eligibility, fixed charges and tariff choice. Neither can be inferred from installed solar capacity alone.
What remains unresolved is the net causal effect of solar expansion on retail rates in cents/kWh and the exact long-run emissions displacement of imported electricity. Even without those estimates, the central research question can be answered: solar expansion was an important part of California’s decarbonization, but high retail prices and import dependence are separate system problems. Progress from here depends as much on storage, transmission, demand flexibility, wildfire and grid-cost control, and tariff design as on adding more solar panels.
Sources and methods
The analysis preserves the original accounting boundary of each source. EIA is used for state price and power-sector emissions comparisons; CEC for total-system supply, imports and planning; CARB for the imported-electricity emissions boundary; CAISO for hourly operating mechanisms; and CPUC for retail rates, wildfire costs, affordability programs and tariff design. Partial-year 2026 data are used only to illustrate current operations, not to replace annual trend evidence.
- U.S. Energy Information Administration: California Electricity Profile 2024
- California Energy Commission: 2024 Total System Electric Generation
- California Air Resources Board: Current California GHG Emission Inventory Data
- California Public Utilities Commission: 2024 California Electric and Gas Utility Costs Report
- California Public Utilities Commission: Electric Rates
- California Public Utilities Commission: 2025 Senate Bill 695 Report
- California Energy Commission: 2021 Total System Electric Generation
- California Energy Commission: 2022 Total System Electric Generation
- California Energy Commission: 2023 Total System Electric Generation
- U.S. Energy Information Administration: United States Electricity Profile 2024
- U.S. Energy Information Administration: Washington Electricity Profile 2024
- U.S. Energy Information Administration: Oregon Electricity Profile 2024
- U.S. Energy Information Administration: Arizona Electricity Profile 2024
- U.S. Energy Information Administration: Nevada Electricity Profile 2024
- California Independent System Operator: Managing the evolving grid
- U.S. Energy Information Administration: Solar generation in CAISO surpassed natural gas in the first five months of 2026
- California Public Utilities Commission: Net Billing Tariff
- California Public Utilities Commission: CARE/FERA Program
- California Energy Commission: Clean Energy Serving California
- California Energy Commission: California Energy Demand, 2025-2045
- California Energy Commission: Low-Income Residential Energy Bill Impact Analysis
- California Public Utilities Commission: Affordability
- California Energy Commission: Behind the Meter Distributed Generation Forecast
- California Energy Commission / CPUC / CARB: 2025 SB 100 Joint Agency Report
Green Map creates custom-edited map images using open geographic data sources such as geoBoundaries, Natural Earth, OpenStreetMap, and government open data. These maps are edited visual materials, not raw data files, and are provided for education, documents, presentations, and graphic reference.