Planet Money (NPR)

Episode: What AI data centers are doing to your electric bill
Date: December 20, 2025
Hosts: Keith Romer, Jeff Guo

Episode Overview

This episode dives into how the explosive growth of AI data centers is influencing Americans’ electricity bills, using a case study in central Ohio. The Planet Money team investigates why residential electric rates are rising rapidly, who’s really to blame, and what’s happening behind the scenes in the power industry. The story follows Ken and Carol Apacki, retirees who have noticed a near-doubling in their electricity rate. The hosts trace the ripple effects of data center expansion from local distribution, to transmission, to the generation and market structure that determines electric prices.

Key Discussion Points & Insights

1. Meet Ken and Carol: The Case of the Rising Electric Bill

• [02:32] Ken Apacki has meticulously tracked his electricity usage and cost since 2020 in a spreadsheet and noticed his rate nearly doubled from 11 cents to 19 cents/kWh over five years.
• [03:45] Ken’s theory: the construction of many new data centers in central Ohio is behind the increase. There are now about 130 data centers in the region.
• [04:11] Carol echoes what they’ve been hearing: rates are going up, the grid is stretched thin.

2. Data Centers’ Physical Impact on the Landscape

• [07:19–09:42] Keith, Ken, and Carol tour the countryside around Granville, Ohio. They see rapid transformations: fields converted to data center construction sites, endless rows of bland, massive buildings, new power lines, and substations.
  • “Every time we come, we see some new vast wasteland.” – Carol Apacki [09:34]
  • "It was just miles and miles of this." – Keith Romer [08:28]

3. The First Layer: Local Distribution & Utilities

• [10:34–15:56] Mark Ryder, President of AEP Ohio, explains:
  • The sheer scale: A new neighborhood might need 1 megawatt, but cluster data centers want 50–3,000 megawatts – the equivalent of an entire city’s demand.
    • “Just one of those bigger data center clusters … would use the same amount of electricity as the entire city of Fort Worth, Texas.” – Jeff Guo [11:32]
  • Utilities used to “socialize” the cost of grid upgrades across everyone. But massive data center requests threatened to burden households with higher rates, especially if projects stalled or failed.
  • AEP Ohio paused all new data center connections in 2023, creating new rules: Data centers now must pay for a bigger share (up to 85%) of capacity they request, post collateral, and prove they’re “real.”
    • “We had to flush out who’s real and who’s not. And you do that with collateral requirements. That’s big money.” – Mark Ryder [15:25]
• Outcome: New rules mean households pay for less of the new infrastructure, but still some. Experts estimate distribution accounts for only 10–20% of the recent price rise.

4. The Second Layer: Transmission

• [16:56–19:44] Kamran Ali, who oversees a multi-state transmission grid, likens the system to an interstate highway with color-coded voltage “lanes.” Massive new demand is driving up costs for regional high-voltage lines.
  • “It took us 100 years to get to 40,000 megawatt. We have currently signed agreements to bring on another 28,000 megawatts by 2030—in a span of five years.” – Kamran Ali [18:50]
• Transmission upgrades, recovered via federal rates, contribute less than 20% of the bill rise—so the bulk of the increase must be elsewhere.

5. The Biggest Layer: Generation & the Capacity Market

• [21:54–32:32] The largest driver is power generation, specifically a “capacity market” system intended to ensure enough supply in the future.
  • Brief history: Ohio broke up vertically-integrated utilities 25 years ago, seeking lower prices through competition.
  • Capacity Market Explained:
    • PJM, the grid operator for 20% of Americans, pays power plants to promise they'll be available in the future—NOT for the actual electricity, but for capacity to generate if ever needed.
    • “Local power distributors have to pay power plant operators just for existing, whether or not those plants ever generate any actual power to sell.” – Keith Romer [27:13]
  • The problem: With data center demand booming, there’s not enough generation being built. Capacity market prices shot up 10x in recent years, costing utilities (and ratepayers) $12 billion more in a single year—mostly paid to existing plants, not new ones.
    • “PJM’s prices for generation went up by a factor of 10.” – Kathy Kunkel [28:53]
    • “$12 billion extra dollars … just to guarantee that there would even be enough electricity to buy in the future.” – Keith Romer [29:01]
  • Why aren’t new power plants being built?
    • The market doesn’t guarantee stable, long-term income (producers fear prices may drop the next year).
    • Renewables like wind/solar count for less in capacity calculations, since they aren’t always “on.”
    • Shortages of key parts (like natural gas turbines); long delays to connect to the grid.
    • “So what used to be like a two, two and a half year wait time to get your turbine has now turned into a five or six year wait time.” – Kathy Kunkel [30:51]
  • So, residential customers are paying more—but they’re mostly rewarding pre-existing power plants, not incentivizing new ones.
    • “Most of the money from the capacity market just ends up going to existing power plants.” – Kathy Kunkel [31:55]
    • “The reason Ken and Carol’s electricity prices went up was mostly … the way the capacity market isn’t really working the way it is supposed to.” – Jeff Guo [32:01]

6. The Systemic Takeaway: Ordinary People Subsidizing Tech Expansion

• [32:51] “I think it's almost inevitable, the way that these structures are set up, that ordinary people are going to end up subsidizing the wealthiest industry in the world.” – Kathy Kunkel
• [33:05–33:30] Demand pressure from data centers is being felt across the U.S. Bills are rising, but policymaker fixes are elusive:
  • “The electricity world, it’s in chaos. Everyone’s got their idea of the best way to fix things ... maybe we should make it easier for new power plants to connect to the grid, or maybe we should stop connecting new data centers altogether until power supply catches up to demand.” – Keith Romer [33:30]

7. What’s Next?

• PJM recently rejected all a dozen reform proposals, promising a new plan soon.
• The Department of Energy may step in at the federal level.
• Meanwhile, new data centers keep getting built—implying further pressure on rates for “the Kens and Carols of the world.”

Notable Quotes & Memorable Moments

• Spreadsheet Life: “Everything ends up in a spreadsheet.” – Carol Apacki [01:29]
• Visual Metaphor: “Every single voltage is a different color ... 765kv is green, because that’s my favorite.” – Kamran Ali [17:10]
• Plain Truth: “I cannot imagine someone designing this system on purpose.” – Kathy Kunkel [23:30]
• Inevitable Cost: “Ordinary people are going to end up subsidizing the wealthiest industry in the world.” – Kathy Kunkel [32:51]
• Policy Deadlock: “PJM … reviewed a dozen different proposals for how to fix its problem, and they rejected all of them. … So it’s anyone’s guess how any of this plays out.” – Jeff Guo [33:48]

Timeline of Key Segments

• [01:15–03:29] Ken & Carol’s data: discovering their escalating electric bill
• [03:45–04:23] Ken’s hypothesis & central Ohio’s data center boom
• [07:04–09:42] Tour of landscape transformation by data centers
• [10:34–15:56] Interview with Mark Ryder (AEP Ohio): distribution, costs and policy adaptation
• [16:56–19:44] Kamran Ali on the scale and cost of expanding transmission
• [21:54–32:32] Capacity markets, PJM, and the broken economics of power generation (with Kathy Kunkel)
• [32:51–34:10] Summary of who pays and what happens next

Final Takeaway

The boom in AI data centers is putting unprecedented pressure on the U.S. power system. The resulting strain is causing electricity prices—most notably in regions like Ohio—to surge. But it’s not just local infrastructure needs; it’s the strange, intricate market design for generating and guaranteeing future electricity supply that’s passing huge costs onto ordinary people. As tech companies race ahead, the system is struggling, and until major reforms happen, households will keep bearing the brunt.

Episode produced by Sam Yellow Horse Kessler, edited by Jess Jiang, fact-checked by Sierra Juarez and Vito Emanuel, engineered by Sina Lofredo. Executive producer: Alex Goldmark.