Basic aluminum steel-reinforced conductor costs twice what it did. Transformers cost more than twice as much. Switchgear has doubled. Gas power plants run two to three times their old price.
Andy Lubershane's answer to whether data centers are raising electricity prices is no, and also yes — and he says the confusion comes from conflating two different questions that have two different answers.
"I have lots of people like my mom and some random friends being like, "Are data centers making my electricity bill go up?" And I had to – I realized that there is not an easy answer to that."
Lubershane is a partner and the head of research at Energy Impact Partners, where he is a frequent guest on Shayle Kann's show; he wrote the piece behind this conversation after fielding that exact question from people who aren't in the energy industry.
I listened to the full episode so you can skip it. 42 minutes of audio, 24 minutes of reading.
Here are the 7 takeaways that matter.
👤 Guest: Andy Lubershane, Partner and Head of Research at Energy Impact Partners
🎙️ Host: Shayle Kann, who invests in early-stage energy companies at Energy Impact Partners and hosts Catalyst
📰 Published: 10 September 2026 on the Catalyst with Shayle Kann feed
🔴 YouTube | 🔗 Episode page | ⏱️ 42 min | ✅ Time saved: 32 min
Key Takeaways
Locally, a new data center usually lowers your bill; globally, data centers are a major reason bills are rising everywhere
The confusion comes from treating those as one question instead of two
A utility's rate is cost divided by kilowatt-hours sold, and a data center adds far more to the denominator than the numerator
That was especially true five years ago, when most grids still had spare capacity
Utilities are now competing to promise rate cuts because of a public-trust problem, not just economics
A Gallup poll found the top three reasons people oppose data centers are water, energy use, and fear their bill will rise
EPRI's research already shows data centers lowering customer bills by about 6% on average
Every major grid component now costs two to three times what it did, because demand is growing faster than the supply chain can build
Two-thirds of all US utility spending is just maintaining and hardening the existing grid, unrelated to data center growth
Electricity is 1% to 1.5% of the average American's income, but 5% to 10% or more for lower-income and fixed-income households
That is where a genuine local rate cut would matter most
Energy efficiency and load flexibility are the two levers Lubershane expects to get renewed attention as prices rise
1. Two Questions, Two Answers
Shayle Kann introduced the piece his Energy Impact Partners colleague had written and asked him to separate what the data actually shows about data centers and electricity prices.
Lubershane said the honest answer depends entirely on which question is being asked. "I realized that there is not an easy answer to that. There's not one simple answer. The answer is nuanced, and it really depends on your perspective." He said the split is between the local impact — what one data center does to the ratepayers in its own utility territory — and the global impact on prices everywhere
Locally, he said, the answer is generally no; globally, the answer is yes
2. Utilities Compete on Rates
Lubershane walked through the mechanics of a utility rate: total cost to serve all customers, divided by total kilowatt-hours sold.
A new data center adds heavily to the denominator, and used to add little to the numerator. "That was especially true, say, five years ago, where in a lot of utility service territories, there was some spare capacity in the system, some headroom on the grid, right?"
That headroom is mostly gone now, which makes the math murkier — but the negotiated deals still protect ratepayers. Because a single data center can add 10% to 30% or more to a mid-sized utility's total load, Lubershane concluded: "So in fact, I think the utilities and the data centers are doing a pretty good job at the local level, when any given new data center comes to town, at making sure that that data center is not going to raise your rates, and in fact should lower your rates a bit, right?"
Kann said the evidence is already showing up in regulatory filings. "We're now seeing the first like announcements from utilities of going to regulators and requesting rate decreases as a result of large loads in their territory," and he expects more of them as utilities move past an earlier, looser period of negotiating with data center customers
Lubershane agreed, and said the trend is now closer to a requirement than a courtesy. "I would say trending towards are going to have to decrease rates locally," because of the surge in public opposition to data center development
EPRI's research already shows a real, if modest, benefit. "And even if you're not controlling for any kind of correlated variables that might bias the analysis, just kind of like looking at that in a very naive way, you would see that there's basically no correlation between more data centers and higher electricity prices. And when you do control for other variables that might bias the analysis, EPRI found that basically, on average, data centers are lowering customer bills a little bit already. So we are seeing it's not a major impact, it's something like 6%, but data centers have had a beneficial impact for consumers up to this point."
3. The Local Fix Misses Trust
Kann pushed on why the public remains hostile even where the arithmetic favors them.
A Gallup poll found water is the top complaint, ahead of energy use and bill fears — which surprised Lubershane. "The number one reason is water consumption, which was kind of surprising to me... I think it's probably a bit of a red herring in most places outside of very sort of water-restricted areas. But the number two and the number three reasons people cite have to do with energy use"
Utilities, data center operators and their regulators share a trust deficit, Kann said. "And there's an additional layer to it, which is that they just don't trust any of the players who are telling them these things, because it's either going to be the data center companies, or the utility, or maybe the politicians."
Lubershane agreed the risk is real: a genuine local benefit could go uncredited. "And if they've heard that this data center was supposed to make their utility bill go down, and their utility bill does not go down, yeah, clearly that will just compound the problem of lack of trust here" — because global pressures are pushing rates up elsewhere at the same time
He referenced Ohio gubernatorial candidate Vivek Ramaswamy's proposal that data centers should make electricity free for local residents, calling it an extreme version of the same underlying idea — that hyperscalers may need to sweeten deals dramatically to win public buy-in
4. The Global Demand Shock
Kann then turned to the global side of the ledger, where Lubershane said the picture flips entirely.
Data centers are the single biggest driver of a demand shock hitting the entire power-equipment supply chain. "It's happening so quickly that it kind of counts as a demand shock in a macroeconomic sense, at least within this sector, right?" Every step of the supply chain — from equipment manufacturing to deployment — is bottlenecked
Most utility spending has nothing to do with data centers at all, and it's getting more expensive too. "Two-thirds of all utility spending across the board in the United States is just for maintenance and hardening of the grid" — a figure Lubershane called out from the Edison Electric Institute — and that baseline spending is rising because of the same supply-chain pressure
The headroom that once cushioned new load is gone almost everywhere, so any new capacity now competes for scarce, expensive equipment rather than tapping spare capacity built decades ago and already partly depreciated
5. Grid Prices Have Doubled
Kann asked him to name names and numbers.
Every component Lubershane checked has roughly doubled or more. "Conductor is twice as expensive, like basic aluminum steel-reinforced conductor. Transformers are two-plus times more expensive, switchgear is two times more expensive, gas power plants are two to three times more expensive, renewables have gotten more expensive to interconnect to the grid"
Renewables had been one of the few things pushing prices down, and that tailwind is fading. Wind and solar power purchase agreements have often been cheaper than the fossil generation they replace, aided by tax credits. "Now those tax credits over the next four or five years will be expiring," alongside rising interconnection costs
Gas turbine prices keep setting new highs. Kann cited combined-cycle gas turbines selling for around $4,000 a kilowatt; Lubershane said he'd recently seen a deal closer to $3,600, one of the higher price points he'd seen. "But yes, yeah, it seems to keep going up."
Labor is a separate, slower-moving constraint: training a qualified electrician or line worker takes years, which Lubershane called the piece he worries about most on a ten-year view even though labor is a small share of total electricity cost today
6. Everyone Wants a Data Center
Given the split verdict — lower rates locally, higher rates globally — Kann drew out the obvious incentive.
The rational move for any individual utility territory is to want as many data centers as possible. "If we're in a data center version of the prisoner's dilemma of sorts, absolutely," Lubershane said. "Like get it all to my territory to lower my electricity bill, for sure"
He floated a real-world test of how far that logic could go. Kann proposed a small utility offering to fast-track a gigawatt-scale data center in exchange for a rate cut calculated to lower every customer's bill by a fixed percentage, and Lubershane said he would like to see someone actually try it
The upside is capped by how small a share of a data center's costs energy actually is. In an earlier piece Lubershane wrote, called "For AI, Energy is Everything and Energy is Nothing," he argued: "But then once you start running, energy is a pretty small share of the total cost of goods sold for that data center, right? It's something like 5% to 10% of the total cost" — which limits how much a hyperscaler can plausibly commit to give away
For lower-income households, even a modest cut would matter more than the national averages suggest. Electricity has run between 1% and 1.5% of average personal income since 2010 and has been slowly declining, but "I will say like where this really matters is for lower-income consumers, fixed-income consumers, for whom electricity could easily be 5%, 10%, maybe even more of their wallet" — so a real local rate cut would free up a meaningful share of their spending
7. What Actually Fixes It
Asked what could relieve the global price pressure, Lubershane said there is no single answer — only a long list of things that all help a little.
His honest answer is "do everything," because every part of the supply chain is constrained at once. He said there is "such a panoply of potential solutions" across the supply chain, utilities and data center operators that it's hard to know where to start
Energy efficiency is the lever he is most hopeful will finally get taken seriously. He said utilities have historically treated efficiency programs as "a regulatory requirement" rather than a planning resource, and rising prices could change that calculus for both consumers and utilities
Load flexibility is the natural next step, matching demand to available supply rather than just shrinking demand. Kann's Energy Impact Partners portfolio company Sparkfund is pursuing a model of "distributed capacity procurement," where utilities pay for batteries on customer properties in exchange for flexibility
High prices are already pulling in new supply-side solutions, from alternative transformer supply chains to solid-state transformers, because those steps respond directly to the price signal — unlike the labor shortage, which Lubershane said will take much longer to fix regardless of price
Bonus Insights
Electricity's share of household spending has been remarkably stable for decades, unlike oil and gas. Kann referenced a chart from a colleague, Nat Bullard, going back to the 1950s showing the share of the average person's income spent on electricity has stayed roughly flat, while the share spent on gasoline has been far more volatile
Kann and Lubershane both described themselves as "DER nerds" who attend the DERVOS distributed-energy-resources conference every year, framing load flexibility and distributed batteries as a longtime interest of theirs that higher prices are now pulling into the mainstream
Lubershane's bottom line is that the local and global stories are both true at once: an individual data center probably lowers the bill of the ratepayers around it, while the aggregate demand it represents is a major force pushing electricity prices up everywhere else — and closing that gap is a decade-long supply-chain and workforce problem, not a single fix.
Products, Companies & Tools Mentioned
Energy Impact Partners (Lubershane and Kann's firm; its portfolio company Sparkfund is pursuing distributed capacity procurement deals that pay for batteries on customer properties)
EPRI (Its research found data centers have, on average, lowered customer bills by about 6% once other variables are controlled for)
ER Grid and Heron Power (Two companies Lubershane cited as responding to record transformer prices — EIP portfolio company ER Grid is building a transformer supply chain, Heron Power makes solid-state transformer alternatives)
Books & Resources Mentioned
"For AI, Energy is Everything and Energy is Nothing" (Lubershane's earlier essay, referenced but not linked in the episode notes, arguing that energy capacity is essential to start a data center but only 5% to 10% of its ongoing operating cost)
The Catalyst episode page (Latitude Media's own notes and links for this episode)
If this was worth your time, send it to someone closer to the industry than you are.
Get the latest market chatter as it happens: