Michael Chua
PowerShift Contributor
First Published Q1 2026
In kitchens and boardrooms across the U.S., the same question keeps surfacing: Why are electricity bills rising?
The numbers explain part of the concern. According to powerlines.org, residential electric bills climbed roughly 40% since 2021. At the same time, utilities sought $31 billion in rate increases for 2025, more than double 2024’s level. The pressure is now shared across the system: customers are frustrated, utilities are managing rising capital needs, and regulators must increasingly weigh massive grid investment needs against the rising political pressure to keep bills affordable.
“The average American has just been stretched,” says Matt Ross, Senior Director of Solutions Engineering at GridX. “It’s not just utility bills. It’s housing, it’s education, it’s childcare. Every piece of that equation matters.”
“People can’t invest in solar panels or heat pumps if they’re struggling to keep the lights on. That’s why I get excited about programs that support vulnerable customers rather than penalize them.”
An Expensive Convergence
Ross points to several overlapping cost drivers. “In context, we have an aging grid, systems are getting to the end of their life, and we’re having to modernize and improve to make the grid more resilient,” he states. “Meanwhile, we have more extreme weather…” magnifying consumer heating and cooling needs and battering infrastructure, prompting utilities to harden their systems.
Into this complex landscape steps the most voracious new consumer of electrons: AI data centers. Hyperscale providers are building megawatt‑scale campuses for machine learning and other high‑performance workloads. Morgan Stanley Institute projects data center load growth of roughly 126 gigawatts each year through 2028, representing nearly one-fifth of global electricity demand growth.
Consumer and policy advocates warn that this expansion could require major new buildouts and additional peaking plants. Further, they note that when utilities build new infrastructure to serve large customers, they typically spread the costs across all ratepayers (Reed, 2025). New Jersey offers a visible example: average electric bills jumped more than 20% in 2025 after regulators approved large transmission expansions (Blum, 2026).
“It’s hard to have both things be true: a grid that is evolving and resilient and not increasing rates,” Ross concedes. “But I don’t think the narrative that utilities are just rate‑basing everything is accurate. The utilities we work with are really diligent.”
Affordability Is a Vibe
For regulators and utilities, affordability is often discussed in terms of energy burden. The National Energy Assistance Directors Association (NEADA) February 2026 Energy Hardship Report found that low‑income households spend an average of 8.6% of income on energy, nearly three times the burden of higher‑income households.
Ross agrees that metrics matter, but argues that the customer experience goes beyond percentages.
“If affordability is a vibe, does it feel like your bill is okay?” he asks. “I have bills that I think are reasonable and bills that aren’t. The average American stays really in tune with whether that trip to the grocery store was more expensive than normal. They know what they like to pay, and when it’s significantly different, it comes as a shock.”
Months of sticker shock have translated into political pressure. Governors in at least six states have introduced measures to cap utilities’ returns or freeze rates, while consumer advocates continue to press that the traditional rate‑of‑return model over-incentivizes capital spending by utilities (Walton, 2026).
Behind the headlines lies a human crisis. Household utility debt climbed from $17.5 billion at the end of 2023 to $23 billion by mid‑2025. About 21 million homes, roughly one in six, are behind on their bills, according to NEADA’s September 2025 Report. Disconnections are also rising, and in much of the country, protections against shutoffs during extreme heat remain limited.
“People can’t invest in solar panels or heat pumps if they’re struggling to keep the lights on,” says Ross. “That’s why I get excited about programs that support vulnerable customers rather than penalize them.”
The Simple Math of Data Center Economics
After more than a decade of working with utilities on rate design and analytics, Ross reaches an unconventional conclusion: hyperscale data centers could actually help make electricity cheaper.
“I don’t think it’s accurate to say data centers are increasing the peakiness of the grid,” he argues. “They’re just elevating the energy consumption.”
His reasoning is straightforward. Utilities earn revenue by charging customers for the energy delivered over wires and poles, which carry mostly fixed costs. “If you run more energy through [the same assets], you’re able to recover more revenue,” Ross explains. “If you’re getting more revenue out of the same infrastructure, that means you have the opportunity to decrease rates while covering the same costs and meeting your return on equity requirements.”
Recent industry research supports part of that view. Deloitte’s 2026 outlook notes that some hyperscalers can flex between 10% and 30% of data center load during peak events without disrupting operations, which could help grid operators manage short-term system stress by responding instantly to grid fluctuations and functioning as batteries. Some states, including Indiana, now require data center developers to fund grid upgrades. These practices may help, rather than hurt, general-consumer affordability.
Still, Ross acknowledges that this opportunity comes with a ticking clock. “If we want to build while keeping affordability down, we need new rates and riders to incentivize peak reduction,” he says. “We need to create carveouts for the most vulnerable customers so their bills don’t go up. And we need to do it fast. We’re in an AI arms race with China. And, China is really good at moving fast. When they decide they want to build electric vehicles, for example, they do it at a scale that’s hard to match.”
Accelerating Rate Design
Traditional rate-case cycles often take years to design, file, litigate, and implement. By the time a new tariff is approved, demand patterns, fuel prices, or load composition may have changed. Ross describes that challenge as one of “velocity,” and argues that both utilities and regulators need faster feedback loops.
“Typical rate design sees yearly general rate cases or every three years,” he says. “That’s not fast enough for responding to these challenges.”
Asked where he has seen affordability and velocity addressed effectively, Ross points to a recent collaboration with a utility in the Midwest.
“Their residential fixed charges had not been increased in a decade, which led to opportunities of inequitable cost recovery among residential customers.” Ross recalls. “In just a three‑month period, we were able to take in all their data and put it to work.”
Because the utility lacked granular socio-economic customer data, Ross’s team used participation in the federal Low Income Home Energy Assistance Program (LIHEAP) as a proxy and ran population‑wide simulations across alternative rate structures.
Regulators initially leaned toward increasing volumetric energy charges. The utility’s preferred path was to raise the fixed charge while keeping per‑kilowatt‑hour prices lower.
“Their data was able to answer it,” Ross says. “The fixed charge mitigated the impact more substantially than increasing the energy charge.”
Turning Loads into Assets
Demand flexibility forms another prong of the velocity agenda. Instead of adding peaker plants that sit idle most of the year, utilities can reduce peak pressure by paying customers to shift or reduce their consumption.
Ross seems particularly bullish on electric vehicles and data center backup systems. “On the hottest day of the year, I do not want to turn off my air conditioner,” he confesses. “I’d much rather sell power back from my electric vehicle. I’m not going anywhere. I just want to sit at home in a nice cool house, and there’s no market for that energy.”
California has launched pilot programs to use EVs as mini power plants, but most U.S. markets still lack tariffs that compensate homeowners for exporting power during peak intervals. Ross argues that data centers could also play a meaningful role in system balancing.
“They all have backup power: gas, diesel, or batteries,” Ross says. “They could use those to effectively curtail load during peak times. You’re talking about a gigawatt of load; if they can curtail one percent, that’s a lot of energy.”
Some hyperscale operators already explore on‑site microgrids and contract curtailment commitments, but broad participation in demand response remains limited. Ross insists that technology is not the barrier. Smart meters can already measure energy flowing in both directions, and grid integration standards keep improving. He argues that tariff structures and compensation mechanisms need to create worthwhile incentives for customers to provide flexibility.
Empowering Customers Through Data
Even with advanced technology, affordability ultimately plays out at the kitchen table. Utilities cannot expect customers to embrace new rate structures or conservation programs if they do not understand their options.
“Customers don’t always know they have choices,” Ross explains. “People underestimate how valuable it is to have access to your usage data in a very easy‑to‑digest visual.”
He speaks from personal experience: when he moved from a GridX‑served utility to one without modern tools, he struggled to assess whether investing in a heat pump penciled out financially.
“I did it on gut feel, which does not feel good,” he recalls.
Ross insists that utilities must invest in data systems that allow them to analyze smart‑meter data quickly and share it securely with vendors. “Have a real modern data strategy,” he urges. “You want to be able to easily but securely share data with best‑of‑breed vendors because it’s going to take a combination of utilities and partners to move the needle. Some utilities are far ahead in how they manage data; others with legacy systems are hampered.”
“Have a real modern data strategy. You want to be able to easily but securely share data with best‑of‑breed vendors because it’s going to take a combination of utilities and partners to move the needle.”
The Road to 2030
What does affordability look like in the coming years? Ross remains cautiously optimistic. By 2030, he believes the industry will know whether running more energy through the system can create downward pressure on rates and whether large customers can be true grid partners. For that to happen, utilities must accelerate the evolution of rate design, embrace demand flexibility, and continue to invest in reliability. Policymakers will need regulatory frameworks that support non‑wires alternatives while protecting vulnerable households. Customers must be empowered with clear, personalized information so they can make decisions about when and how they use electricity.
Major uncertainties remain. Natural gas markets could stay volatile. Geopolitical disruptions could reshape equipment supply chains. Climate stress could intensify infrastructure risk. AI-driven load growth could exceed even aggressive projections.
“No one can perfectly predict the future,” Ross notes. “That’s why we run multiple scenarios to understand the range of possibilities: the hottest summer on record, the coldest winter on record, and everything in between.”
Ultimately, velocity involves more than speed; it’s about alignment. It involves aligning infrastructure investment with demand, aligning rate design with equitable outcomes, and aligning modern technology with customers’ everyday experience. If utilities, regulators, and their partners achieve that alignment, then the next decade of electrification and AI could be defined not by sticker shock but by shared prosperity.
Receive the Magazine today.
