EPRI Study Finds Downward Rate Pressure
Asa Watten is an EPRI economist. Geoff Blanford is EPRI principal technical executive.
Rising electricity prices have been linked by many observers to the rapid growth of data centers. However, a new study from EPRI reaches a different conclusion. Between 2015 and 2024, data center growth in the United States modestly reduced average retail electricity rates, compared to what they would have been, rather than increasing them.
Other studies (such as, Wiser et al., 2025 and EPRI’s Win-Win Watts report) show that electricity prices have tended to grow more slowly in states that saw higher load growth. This research is the first to estimate the price impact of new load from data centers using a causal methodology. The study concludes that, on average, data centers put downward pressure on residential electricity rates during the period analyzed.
Data Centers Not Correlated with Higher Electricity Rates
Data centers are becoming a major part of the U.S. energy landscape. As artificial intelligence and other digital services continue to expand, data centers consumed roughly 4.5 percent of U.S. electricity in 2024 and are projected to account for 9 percent to 17 percent of national electricity demand by 2030.
Data center load growth is concentrated in certain areas, and while other drivers of load growth are at play, electricity demand hasn’t increased everywhere. Comparing price changes to demand changes across states, areas with stronger load growth often experienced lower price increases, and vice versa.
In Virginia, the largest data center market in the U.S., data centers consume more than 20 percent of the state’s electricity. Yet Virginia experienced lower-than-average increases in retail electricity rates during recent years. Meanwhile, California saw average electricity prices increase by nearly 40 percent between 2019 and 2024 despite modest data center growth, primarily driven instead by wildfire-related utility costs.
However, these observations only indicate correlation, not causation. Data centers may have been built in places where electricity was already inexpensive. To resolve this question, we used an econometric approach that removes the influence of location-selection effects and isolates the impact of data center growth on retail prices.
Doubling Data Center Capacity Reduced Residential Prices by About 4 Percent
The study’s headline finding is that a doubling of data center capacity caused residential retail electricity prices to be approximately 4 percent lower than they would have been otherwise during the study period of 2015 and 2024. The average customer lived in a state where data center capacity increased by roughly 160 percent from 2019 to 2024. Based on the study’s results, average residential electricity prices would have been about 6 percent higher without data center growth.
Our analysis did not find evidence of systematic cost shifting from industrial and commercial customers to residential customers. Instead, the data suggest that growth in demand tends to reduce average costs both within and between customer classes.
Why More Demand Can Lead to Lower Prices
At first glance, the findings may seem counterintuitive. Conventional supply-and-demand logic suggests that higher demand should lead to higher prices. The electricity sector, however, operates differently from many other markets.
Power systems have large, fixed costs embedded in generation, transmission lines, substations, and distribution infrastructure. When durable new demand arrives, utilities can spread those fixed costs across a larger volume of customers, lowering average costs per kilowatt-hour. Sustained demand growth also encourages new investments, which often have lower average costs than existing assets due to improvements in technology cost and performance over time.
Data Analysis Challenges the Popular Narrative
The study’s findings conflict with the widespread belief that data centers are driving electricity bills higher. Retail electricity rates have increased sharply in recent years, but the main drivers have been economy-wide inflation and higher distribution costs from grid modernization and resilience. Our econometric analysis of the data shows that load growth and data center additions are not among the causal drivers.
Implications for Electrification
The findings may have implications beyond data centers. The analysis finds similar economic dynamics can apply to broader electrification efforts, including the adoption of electric vehicles and heat pumps.
Like data centers, these technologies increase electricity demand. If that demand is durable and the power system can continue adding new capacity efficiently (and if demand peaks are managed well), the resulting growth may help spread fixed infrastructure costs and support additional investment in lower-cost generation resources. Under those conditions, electrification need not increase electricity rates and may even help lower them while reducing emissions and improving affordability for consumers.
Important Caveats
Our analysis cautions that this outcome depends on the industry’s ability to continue building new infrastructure. Supply-chain constraints, equipment shortages, permitting delays, and other barriers could weaken or reverse the historical relationship observed in the study.
Demand flexibility from data centers and other loads will likely play a key role in ensuring system adequacy during this period of rapid growth. Uncertainty is another factor, as utilities invest based on expected demand. If capacity is added to serve projected load that fails to materialize, costs could be spread over fewer kilowatt-hours, increasing rates.
The Bottom Line
Our analysis challenged one of the most common assumptions in today’s energy debate and found data center growth had a modest negative impact on retail rates. The mechanism is well understood and well-grounded in economic theory: durable new demand can spread fixed costs across more usage and pulls in cheaper new capacity.
The broader lesson is that growth in demand — whether from data centers, electric vehicles, or heat pumps—doesn’t necessarily translate into higher electric bills. It can, in fact, lower them.
To learn more about the analysis, visit: Have Data Centers Raised Your Electric Bill? Causal Evidence from the United States. Summary and key insights are available here.
