Why electricity prices in California are so high

Electricity prices are rising worldwide, and no one seems to agree on why. Common theories include increased use of intermittent renewables, rising demand, and higher natural gas prices. It’s difficult to generalize about the problem since electricity generation and delivery are complex and differ across regions. The best way to understand the problem is to analyze specific areas in depth. In this analysis, we will focus on California for 2024, which is often portrayed as a model (or warning) for how a large region can decarbonize its grid.

About three-quarters of Californians obtain their power from Investor Owned Utilities (IOUs) regulated by the California Public Utilities Commission (CPUC). The rest are served by Publicly Owned Utilities (POUs), which are vertically integrated and self-governed. This analysis focuses on the IOUs, as they supply the majority of the state’s load and are subject to the most transparent reporting requirements.

Nearly all of the IOU load is served by three large utilities: Pacific Gas & Electric (PG&E), Southern California Edison (SCE), and San Diego Gas & Electric (SDG&E). A handful of small IOUs serve the remaining share. IOUs are regulated by the CPUC, and their transmission grids in most of their service territory are operated by the California Independent System Operator (CAISO). Most of the data in this report comes from CPUC filings and legislative reports for 2024, published in 2025. This is the most recent data publicly available.

IOUs are responsible for delivering electricity, maintaining the distribution and transmission infrastructure, and billing customers. While these utilities own some generation, more power is purchased through contracts with independent generators. Over the past decade, much of the IOU customer base has exercised its option to switch to Community Choice Aggregators (CCAs) to provide the generation portion of their bill. These are public agencies formed by cities and counties. In addition, Direct Access (DA) programs are an option for large commercial customers. Even if customers opt to get generation from a CCA or DA, the IOU is still responsible for the remainder of the service.

As of 2024, about half of the generation is provided by an IOU, and the rest is from a CCA or DA provider. Customers choosing a CCA or DA must pay a fee to cover legacy costs from investments the IOUs made, primarily in renewable energy, when prices were higher than they are today. See Figure 1.

.Figure 1. The layout of how electricity is generated and delivered in California.

High-level view of cost

In this report, we will focus on the all-sector electricity rate. This is the average price for industrial, commercial, and residential consumers. It should not be confused with the residential rate, which is higher. In 2024, the all-sector rate in the US was 13.48 ¢/kWh; in California, it was 29.07 ¢/kWh. In the US, the residential rate was 16.60 ¢/kWh, while in California, it was 31.22 ¢/kWh. The all-sector rate of California’s IOUs is slightly higher, at 30.98 ¢/kWh and 35.51 ¢/kWh for residential customers. The all-sector rate of California’s POUs is lower, at ~23.6 ¢/kWh.

Figure 2 shows the historical all-sector rate for both the US and California. California historically paid a small premium for electricity, but rates have diverged over the last decade. The US rate tracked inflation until 2020, with an uptick beginning shortly after. In California, prices diverged in 2012, then skyrocketed in 2020. What changed?

Figure 2. Historical all-sector rate ¢/kWh.

The pain for consumers is expected to get worse. The CPUC ordered the IOUs to forecast future prices. Figure 3 shows the results of their modeling of residential rates. Prices are predicted to surpass 40 ¢/kWh by 2028. Bundled are customers who receive generation from their utility.

Figure 3. Forecasted rises in residential rates.

Figure 4 depicts the cost components in ¢/kWh for the all-sectors rate. Column 3 shows costs in ¢/kWh of each component, and column 4 shows the percentage of the total cost. The data is derived from these two sources published by the CPUC: link1 and link2.

Figure 4. Cost components in ¢/kWh, average across all sectors, for 2024.

The components of cost tell us where the money is spent, but they can’t tell us why the costs are so high. To understand that, we must take a closer look at each of the major categories.

Generation

We begin with a focus on the generation piece for customers opting to buy generation from their IOU (bundled). We will assume CCAs and DAs have similar costs due to the competition for generation customers. The total unadjusted revenue requirement for generation in 2024 was $11.3 billion. The revenue requirement is the amount of money a utility is allowed to collect from ratepayers. The all-sector generation cost alone was 13.12 ¢/kWh, nearly equal to the US average cost for complete service at 13.48 ¢/kWh.

We already have a serious problem. The generation cost should be closer to 6 ¢/kWh to match the US. Generation costs typically account for about 45% of an electric bill. Figure 5 shows the cost of each component.

Figure 5. All-sector cost components generation in 2024.

Line 6 indicates that most of the cost is attributed to purchasing power from privately owned power plants, utilizing contracts known as Power Purchase Agreements (PPAs). Line 7 shows the price of fuel for the utility-generated power, primarily natural gas. Line 8 shows the capital and maintenance costs of utility-owned generation. Line 10 shows the return to investors. You could call this the utility’s profits. They receive about a 10% return on the money they invest in new projects. There are no obvious red flags, except the excessive cost.

Next, we check the wholesale prices. Figure 6 shows the weighted average of wholesale prices for all nodes in the state. In 2024, the cost was 4.3 ¢/kWh, which is about one-third of the generation cost. Note that the spike in 2022 did not result in a corresponding spike in electricity prices, as shown in Figure 2.

Figure 6. California statewide weighted average wholesale prices.

Renewable energy advocates often use low wholesale prices as evidence that a high penetration of intermittent renewable energy does not lead to higher electricity prices. In fact, a high penetration of renewable energy typically suppresses wholesale prices, but not because it is cheaper. Renewable energy is rarely bid into the wholesale market; instead, it is self-scheduled to run regardless of the clearing price. This happens for a variety of reasons: policy to lower CO₂ emissions, it has zero marginal costs, and it has guaranteed prices via a fixed-price Power Purchase Agreement (PPA).

Self-scheduling has the effect of bidding zero. Gas plants come in last to fill in any missing energy for the time slot. The gas plants then determine the clearing price that all bidders and self-schedulers receive. It is relatively common for the clearing price to be zero, which squeezes gas plants out, but the renewable plants are still paid according to their PPA contracts.

California’s Load Serving Entities (LSEs) depend very little on the wholesale market. LSE is the regulatory term used to describe all organizations that provide consumers with electricity: utilities, CCAs, and DAs. They typically purchase power from privately owned power plants using a PPA and rely on their own assets. The wholesale market is primarily used for final load balancing; it thus plays a minor role in determining final electricity prices.

The three distinct components that determine the cost of electricity:

Cost of electricity = Fixed costs + Variable costs + Legacy costs

The goal is to satisfy consumer demand without blackouts. Having an abundance of energy alone cannot accomplish this. The entire electrical system must be designed to handle the worst-case scenario: generation, transmission, and distribution. This typically occurs on a hot summer evening when people return home from work and the air conditioners are running at full blast. The fixed costs are associated with building and maintaining the entire system, regardless of how frequently the enabling assets are used. The whole system must have sufficient capacity to handle the worst-case scenario.

The variable costs are the cost of fuel and wear on generators when the equipment is in use. In addition, California made significant investments in solar energy back when it was more expensive; these legacy costs still factor into current prices. The fixed cost accounts for over 90% of the costs. This fact is one of the most essential concepts for understanding the cost dynamics of electricity!

The most difficult task in providing reliable electricity is ensuring that peak demand is always met. To guarantee this, California depends on a little-discussed market called Resource Adequacy (RA). This is analogous to a capacity market, a common tool used by many other states to ensure reliability. The CPUC determines the peak demand for each month. Every LSE is required to have sufficient reliable capacity to meet these peaks, plus a reserve margin (currently 17%). The keyword here is reliable.

The CPUC assigns a reliability rating to each source of energy. This is called the Effective Load Carrying Capacity (ELCC). For example, an ELCC of 5% means that only 5% of the nameplate capacity can be counted on to meet peak demand. Solar plants have the lowest ELCC, while dispatchable plants have the highest. ELCCs are calculated using computer simulations. Each LSE meets its RA obligation with a combination of generators that it owns, PPAs, and by purchasing RA from the RA market.

RA is priced in $/kW-month. A kW-month of RA is a kW of capacity that can be counted on to meet peak demand during the month it was purchased for. An RA asset may or may not be in use during that month, but it must be prepared to provide electricity if needed. The ELCC of an asset is used to determine the percentage of its nameplate capacity that is counted.

The price of RA has been rising dramatically in recent years. Figure 7 illustrates the exponential growth. RA cost rises as the percentage of intermittent energy increases.

Figure 7. Weighted Average Price of System RA ($/kW-month), 2018-2023.

The RA obligation is achieved by deploying a mixture of RA assets:

1. Dispatchable power plant: peakers, combined cycle gas, nuclear, hydro.

2. Storage. Mostly lithium-ion batteries.

3. Demand response. Consumers are paid not to consume during consumption peaks.

Lithium-ion storage is playing a significant role in higher generation costs. As of April 2025, California has deployed 13,248 MW of utility-scale lithium-ion battery capacity. Using 2024 EIA prices, the overnight cost of this is $21 billion.

A key point to remember is that dispatchable plants can generate energy and simultaneously supply RA. A solar or wind plant generates energy, but must be backed up with RA at an additional cost.

Cost of dispatchable generation = Energy

Cost of intermittent generation = Energy + RA

If we use the procedure outlined by the CPUC for calculating RA costs, the value of RA for the IOUs in 2024 is approximately $7 billion. This compares to the total revenue requirement for generation at 11.3 billion. The utilities don’t actually pay this much because they still have a substantial quantity of dispatchable power that requires no additional RA obligation. This dispatchable power can be used to provide RA for their intermittent procurement, and even to provide RA to other LSEs by selling into the RA market. However, the immense value of RA serves as a warning of how expensive achieving reliability is for a grid dominated by intermittent generators.

Figure 8 shows the cost IOUs paid for RPS and non-RPS generation as of 2024. RPS stands for Renewable Portfolio Standard. RPS includes solar, wind, small hydro, biogas, and biomass. Solar and wind account for the vast majority. Although nuclear power and large hydro are CO₂-free, they are not considered RPS-eligible. This designation is needed because California law requires LSEs to have defined levels of RPS energy by specific dates. The non-RPS energy is nuclear, large hydro, and natural gas generation.

Figure 8. Large Investor-Owned Utilities’ 2024 Average RPS and Non-RPS (¢/kWh).

Figure 8 shows that the average cost of RPS-eligible energy is 42% more expensive than non-RPS-eligible energy. Once RA obligations are factored in, the difference is greater yet. Remember, most RPS energy still depends on non-RPS energy to meet its RA requirements. As a result, fossil fuel plants are producing less energy but must be available to run as solar power fades into the evening. Instead of being paid to generate energy, they are paid to be ready to do so. This service is not free.

RPS energy includes older, more expensive solar farms that can be procured at a lower cost today. The legacy cost is the difference between the original cost and the current cost of comparable assets. Legacy costs are often cited as a primary reason for California’s higher electricity prices. The CPUC estimates that the legacy cost differential for the IOUs in 2024 was $1.2 billion, which equates to 1.39 ¢/kWh. If we subtract the legacy cost from the total generation cost, the difference is still almost twice as expensive as the national average. Thus, legacy costs play a minor role in explaining high electricity prices. Moreover, legacy costs are decreasing annually due to the expiration of older PPA contracts.

Figure 9 shows the cost of the utility-procured RPS-eligible power sources. This is a combination of long-standing PPAs and utility-owned. Line 10 shows that PG&E paid 23.5 ¢/kWh for its photovoltaic solar. IOUs made significant investments in solar technology when it was still in its infancy, and as a result, its customers paid higher rates.

Figure 9. Cost of utility-owned RPS ¢/kWh generation in 2024.

Also note the high cost of various forms of smaller-scale renewable energy: biogas, biomass, and small hydro. These add to the diversity of the renewable portfolio but are pretty expensive due to their lack of scale.

Figure 10 shows the historical cost of RPS-eligible energy. It reached a low point of around 3 ¢/kWh in 2021 and has risen steadily ever since. The price in 2024 was 8.1 ¢/kWh for new contracts, double the historical cost of gas, and still requires additional RA expenses.

Figure 10. Historical cost of RPS energy.

Below is the explanation given by the CPUC for the trend in higher RPS prices (primarily solar power):

“2024 conditions appear to have led to an increase in RPS prices: The average price of RPS contracts executed in 2024 was 8.1 ¢/kWh, compared to the 5.9 ¢/kWh average price for contracts executed in 2023 in real dollars. This 37 percent increase was likely driven by an increase in demand due to retail sellers need to meet end of RPS Compliance Period 2021-2024 requirements, uncertainty of supply chain constraints, anticipation of potential continuing increase in the overall levels of inflation, and higher interest rates.”

According to Lazard, the current borrowing rate for a new solar or wind plant is 7.7%. At this rate, interest accounts for about 50% of the levelized cost. Figure 11 shows that the federal funds rate fell to near zero from 2010 to 2022. This created artificially low prices for capital-intensive projects during that period, which are unlikely to be seen again.

Figure 11. Historical federal funds rate.

The claim that the cost of solar energy in the US is in continuous decline is a myth. This myth is based on the steady decrease in the global spot market price of Chinese-made solar modules. However, if we consider the entire system cost for a working solar farm installed in the US, the overnight costs have been stagnating. While the cost of modules has decreased, the cost of other components, such as transformers, has increased. When factoring in rising interest rates, the procurement cost has been growing, as shown in Figure 10. If not for the federal subsidies, the 2024 RPS cost for new contracts would be closer to 11.4 ¢/kWh, with significant RA costs on top of that.

In summary, California IOUs paid an excessive price for a variety of small-scale renewables and first-generation solar power. This led to higher generation costs. However, since the total percentage of the generation mix was fairly low, the RA costs were low as well. As solar power deployment expanded, the price decreased. But as the percentage increased, so did the RA costs. The result is that the cost of generation remains high.

Due to higher interest rates and more realistic development costs, LSEs are currently paying 8.1 ¢/kWh for new solar power contracts. And they still incur additional costs for underutilized fossil fuel plants, batteries, and demand response measures to make them reliable. In contrast, the price for non-RPS energy IOUs paid is 7.1 ¢/kWh, requires no additional RA, and is close to the US average for generation at ~6 ¢/kWh. The 1.1 ¢/kWh difference can be largely explained by the premium price California pays for natural gas, due to various environmental policies.

It is clear from the data that the predominant cause for California’s higher generation costs compared to the national average is its large deployment of intermittent renewable energy and other expensive small-scale forms of renewables. This cost will continue to rise as California moves to higher percentages. The RA costs shown in Figure 7 will continue to rise as more storage is required, and the utilization rate of natural gas plants decreases. Additionally, the termination of federal subsidies (PTC, ITC) in 2027 will result in another 30% to 40% increase in the costs of new solar and wind farms.

Transmission

The transmission grid comprises high-voltage power lines that transmit energy from power plants to substations just outside urban areas. The revenue requirement for the IOUs in 2024 was $4.68 billion. The cost of transmission was 2.71 ¢/kWh. Transmission is typically 10% to 12% of a bill, so we should expect the price to be ~1.48 ¢/kWh compared to the US baseline. California is thus paying a premium of ~1.23 ¢/kWh.

Figure 12 shows the components. Wildfire mitigation costs are sometimes associated with transmission, but Figure 12 shows that this is not the case; as we shall see, this is a major problem in the distribution grid.

Figure 12. Components of cost.

Figure 13 shows that transmission costs have been steadily increasing over the past decade.

Figure 13. The cost of transmission is rising (2014 to 2024).

According to CPUC, the reasons include:

“Historically, much of the increase in the utilities’ revenue requirements was due to capital investments in transmission infrastructure. Significant portions of the IOUs’ transmission rate bases include CAISO approved reliability projects, as well as policy projects needed for meeting clean energy mandates. These CAISO-approved projects expand capacity of the grid, enabling interconnection of new electric generation, as well as compliance with North American Electric Reliability Corporation (NERC) requirements.”

Due to California’s rugged terrain (mountains, deserts, and valleys), it is more expensive to transport power from remote areas suitable for wind and solar plants into cities, compared to traditional power generation, which is closer to load centers. Additionally, California has strict environmental review requirements, making it difficult and time-consuming to build new transmission lines. This will be a challenge in the future.

Distribution

The distribution grid consists of lower-voltage power lines that carry energy from substations located outside urban areas to the millions of homes and businesses. The revenue requirement for the IOUs in 2024 was $22.7 billion. The all-sector cost was 13.16 ¢/kWh. Distribution costs are typically about 45% of a bill, so we should expect the price to be ~6 ¢/kWh, compared to our US baseline. Californians are thus paying a premium of ~7.16¢/kWh. Figure 14 breaks down the component costs.

Figure 14. Cost components of the distribution grid.

The IOUs spent $7.7 billion in 2024 on wildfire mitigation, which equates to 4.2 ¢/kWh. Due to years of fire suppression, exacerbated by warming trends, and the desire to build homes in wooded areas, wildfires caused by downed distribution lines are a never-ending problem. To resolve the situation, the utilities are burying the power lines, but the cost is enormous.

Wildfire mitigation accounts for just over half of the price premium. Other distribution costs are also growing at a rapid rate. According to the CPUC, the annual growth rate for non-wildfire expenses is 11%.

California has made significant investments in the distribution grid to accommodate rooftop solar. The distribution was designed to carry power in one direction from the substation to the home. With the massive adoption of rooftop solar, power now has to flow in both directions. Not only does power have to flow backward, but power coming out of the home can exceed the power going in. This can happen on a sunny day when the rooftop solar system produces multiples of what a home can consume. Accommodating this involves upgrading transformers and reconductoring lines.

In addition, California law requires the rooftop solar systems to use smart inverters. The grid operator or utility must be able to control these to curtail the solar output as needed. This requires a significant investment in hardware capable of managing these devices.

Public purpose programs & Other

California uses the electric bill as a kind of tax to create a hodgepodge of programs designed to promote energy conservation, distributed generation, and to help low-income customers pay their electric bills. Since none of these involve the generation of electricity, these costs add directly to the bill.

Other is an uncategorized set of miscellaneous items: taxes, fees, and credits. Included in this is the cap-and-trade program. Consumers pay a carbon tax for CO₂ emitted by their LSE; the funds are then redistributed back to all ratepayers. The goal is to create a financial incentive to reduce emissions.

Combined, these add 1.99 ¢/kWh to the bill.

Figure 15.

Rooftop solar

On the surface, rooftop solar seems like such a great idea, but this one item accounts for the single largest increase in cost compared to the US average. A high adoption rate makes generation, transmission, and distribution costs all the more expensive. The CPUC estimates that in 2024 $7 billion to $8.5 billion in costs were shifted to consumers without rooftop solar. This equates to a rise in rates of 4 ¢/kWh to 4.6 ¢/kWh.

We previously discussed upgrades to the distribution to accommodate rooftop solar, but there are two additional crucial factors: Net Energy Metering (NEM) and fixed cost avoidance. Each of these accounts for ~50% of the estimated cost shift.

NEM allows rooftop solar customers to sell their excess solar power back to the utility at the retail price. This means that the utility is paying, for instance, ~35 ¢/kWh for electricity that it can buy for ~8 ¢/kWh. The utility depends on its markup to fund its operations; therefore, it now has to raise everyone’s rates to compensate for the lost revenue. California has discontinued NEM for new customers, but it grandfathered it for preexisting customers for 20 years.

In the generation section, we discussed how fixed costs account for over 90% of the costs of providing electricity, and that these costs are determined by the quantity of hardware and labor required to meet peak demand. This includes generators, transmission, and distribution. Yet retail customers are billed based on their level of consumption. This type of billing creates perverse incentives, with rooftop solar being the worst example.

Let’s say rooftop solar customers generate half of their electricity from rooftop solar during the day when the sun is shining, and then smoothly transition to the grid in the evening when the solar power is gone. They end up cutting their bill in half. Sounds great. The problem is that they are not paying for their fair share of the fixed costs. They only saved the utility on fuel that wasn’t consumed, but this is a trivial saving. Even worse, they will depend on the grid in the evening during peak demand for a price less than the actual cost to produce it. A fair price during peak demand, using volumetric billing, would be $1 to $2 per kWh (or more). The utility overcharges during the day, when rooftop solar customers are not contributing, to keep prices lower during the peak.

A fairer way to bill would have a demand charge and a variable charge. The demand charge would be set based on the maximum consumption. The variable charge would be the fuel consumed and the wear on the equipment. This is basically how industrial customers are billed. Billing like this would solve the fixed cost avoidance problem. It would also incentivize the use of home batteries and conservation during peak demand periods without the need for complex programs. However, it would disincentivize the use of rooftop solar. In the US, utilities can produce solar energy at about one-third of rooftop installations; it’s hard to see the logic in incentivizing rooftop solar.

POUs versus IOUs

California’s POUs provide service for ~20% less than its IOUs. This has raised questions about the value of having IOUs. Here are some reasons why POUs have lower prices:

1. POUs serve large urban areas like LA. The higher density cuts the length of transmission and distribution lines per GWh delivered by ~50%.

2. POUs serving large cities have a fraction of the wildfire mitigation costs.

3. POUs have a much lower rooftop solar adoption rate due to a high percentage of multifamily housing and shading from tall buildings.

4. POUs spend less on public purpose programs.

5. POUs can fund new projects at lower interest rates with tax-free municipal bonds.

One disadvantage of a POU is that the municipality assumes all the risk associated with owning such an expensive asset. While rare, publicly owned utilities have gone bankrupt, causing great economic hardship for the people.

Why some states with renewables have lower prices

Globally, there is a strong correlation between regions with high percentages of intermittent renewable energy and higher electricity prices. However, there are some notable exceptions. In the US, the Great Plains region has intermittent energy consumption levels of 35% to 40%, and all-sector rates are still slightly lower than the US average. Here are some of the reasons for their superior results.

1. The Great Plains region has some of the strongest onshore wind speeds in the world. Wind farms in this region produce ~50% more energy than those in California and other coastal states.

2. The higher capacity factor for wind farms in these states results in lower RA costs due to their higher ELCC.

3. The land is flat with a low population density. This enables the construction of huge contiguous plants with lower transmission costs. The average wind farm in Texas is ~3 times the size of one in California. A typical Texas solar farm is ~4 times the size of one in California. The larger a plant is, the cheaper the unit cost for construction and operation, regardless of the source: solar, wind, storage, gas, hydro, or nuclear.

4. These states avoided other non-generation related costs incurred in California: polices pushing rooftop solar, wildfire mitigation, public purpose programs, legacy costs, and excess regulation.

5. These states are a Mecca for Renewable Energy Credits (RECs) and Virtual Power Purchase Agreements that shift costs to wealthy corporations in other regions. The nation’s big tech companies use these financial tools to claim they are CO₂-free, and they pay a hefty price. In California, LSEs must retire their RECs to comply with the state-mandated Renewal Portfolio Standard (RPS); thus, they have no economic value. In California, PPAs are not funded by wealthy companies, but by local LSEs, and the costs are passed on to consumers.

These details matter. Policy makers must remind themselves that a solar farm in West Texas will be a fraction of the cost of one in New York state, with its rugged landscapes and brutal winters. let’s not forget that wind and solar energy are heavily subsidized by the federal government until 2027, and these subsidies played a significant role in keeping costs down.

Even with all their advantages, the Great Plains states are starting to face their own challenges. Texas just passed a law allowing ERCOT to curtail large consumers at a moment’s notice. A good rule of thumb is that when the intermittent renewable penetration is less than or equal to the capacity factor of the intermittent source, curtailment and congestion will be low without the need for storage. These states are just starting to cross that threshold. Prices at 35% intermittent are significantly cheaper than those at 50%, and prices at 50% are significantly cheaper than at 65%, and so on.

Conclusion

California’s high electricity rates are not due to bad luck or mysterious forces that no one can understand. Except for wildfire mitigation, they are the result of policy choices aimed at advancing an energy transition.

The plan is to decarbonize the grid, then electrify everything. But if the transition drives up electricity prices, it will likely fail. The grid is responsible for ~30% of CO₂ emissions; thus, reductions there have only a modest effect. To electrify everything, we must electrify heating, transportation, and industry.

Operating a heat pump is cost-competitive with a gas furnace at ~18 ¢/kWh. The cost of driving an EV is the same as a hybrid at ~27 ¢/kWh. The cost of high-speed charging in California is now around twice that or more. So, consumers are being asked to pay double the cost of heating and driving, even as electricity prices continue to rise. Industry is the most sensitive of all to high energy prices. Electrifying industry is a challenge with industrial prices in the US averaging ~7 ¢/kWh; in California, they are currently triple that.

California is an incredibly wealthy state with exceptional weather. This has enabled them to cope with these high prices, but California is more a warning than a role model for the rest of the world. For an energy transition to succeed, the cost of electricity must be cheap.