NPR.
Data centers are getting a lot of heat right now.
There are protests against them, cities voting to keep them out.
You even have both Bernie Sanders and Ron DeSantis speaking out against data centers.
Two men I'm pretty sure wouldn't even agree that, you know, water is wet.
You know data centers.
They are essentially warehouses packed with computers that make cloud computing and AI possible.
And there are a lot of different reasons people have for opposing them, like concerns about AI and worries about the tons of water they use.
And then there's the fear that data centers will lead to higher electric bills.
Electricity prices are up about 7% year over year, way higher than overall inflation.
And that's driven in part by this flurry of power-hungry data centers.
And the concern is that all the new centers coming online will only drive electric rates even higher.
But that is not inevitable.
This data center electric bill upcharge is not a guarantee.
In fact, it is even possible for data centers to cause power bills to go down.
This is The Indicator from Planet Money.
I'm Waylon Wong.
And I'm Stephen Basaja.
Today on the show, we look at three possible futures for your power bill.
Futures where power companies over-prepare for AI, under-prepare in the Goldilocks scenario where everything goes right and your electric bill could actually get cheaper.
This message comes from LinkedIn ads.
One of the hardest parts about B2B marketing is reaching the right audience.
That's why you need LinkedIn ads.
You can target your buyers by job title, company, role, seniority, and skills.
All the professionals you need to reach in one place.
Get a $250 credit on your next campaign so you can try it yourself.
Just go to LinkedIn.com slash NPR pod.
That's LinkedIn.com slash NPR pod.
Terms and conditions apply only on LinkedIn ads.
So your people can focus on the work that they want to do.
That's putting AI agents to work for people.
It's your turn.
Get started at servicenow.com slash AI dash agents.
To lay out our three futures is Greg Upton.
He's the executive director for Louisiana State University's Center for Energy Studies.
There's no reason that inflation-adjusted electricity prices have to increase over time.
It sounds like you're not a electric price data center doomsayer.
I'm definitely not a doomsayer.
The groups responsible for determining which future we end up with are power companies and their regulators.
The regulator could be called the Public Service Commission, the Public Utility Commission.
The name can change based upon the state.
Whatever you call them, those regulators and utilities need to get together and make a bet.
The bet being on just how much electricity data centers will actually need.
And there is a huge range of predictions here.
One estimate is that data centers will take up an incredible 16 of all energy used in the US by 2030.
I mean, huge amounts.
Some other estimates, though, are way more conservative.
Like one is less than 7%.
Greg says.
Essentially, the utilities present regulators with a story about how much demand they think there will be from the new data center and how much power the utility wants to bring online to match, like by building new natural gas plants, renewable energy sources and nuclear power plants.
And then the regulator, if they have a successful story, is going to approve the utility in order to build that additional generation.
And then we'll have to wait to see if they bet right.
And that will likely play out in one of three scenarios.
Scenario one, they got it wrong.
Data centers don't need that much electricity.
Maybe AI flops or data centers just get much more energy efficient.
Whatever the reason, utilities have overbuilt and have excess power.
Greg says utilities tend to end up overbuilding.
Their version of being cautious probably doesn't hurt that they, you know, make money by building more plants.
And there are still benefits in this construction zealous future.
Well, on one hand, the whole grid is actually going to run more efficiently.
You've got this new generator.
That new generator is probably going to be more efficient than a lot of your legacy older generators.
But there's also a big downside.
On the other hand, you built this capital.
And that capital is going to have to be paid for by ratepayers.
You might think all that extra supply of power with lower costs, but it's hard to store electricity.
Batteries just aren't that good yet.
So we're stuck with paying for that unused power or power plants just sitting there idle.
And so if you build over bill generation, you might have fuel savings in the short run.
But the amount of the capital that you're spending is not worth.
Regulators can hedge against this by requiring data centers to pay for a certain amount of power, even if they don't use it all.
That's what Ohio regulators did.
Data centers there will have to pay at least 85% of electricity they said they would need.
Greg says this overbilled scenario would likely cause residents' power bills in the community to go up.
That's scenario one.
The other scenario is that... electricity demand grows very rapidly, way faster than you projected.
So this is the case where the utilities underbuilt.
They're not generating enough power to meet the new demand.
Utilities could then start building new plants, but that takes time.
So, while they try to catch up, Greg says they'd be bringing online older power plants, plants that are less efficient and cost more to run, and that drives up the price of electricity.
And so in the instance that you don't build out the new generation, in that instance you can also have higher costs.
Then there's scenario three, building not too much, not too little, but getting it just right.
And so it's this Goldilocks zone.
This is that magical scenario where electric bills could be lower or, at least you know, rise slower than inflation.
And it comes down to economies of scale.
Essentially, economies of scale means that companies often get more efficient as their business grows.
The same concept applies to power generation.
Economies of scale definitely definitely can play in in order to supply that in the most efficient way.
Newer, larger plants can be more cost efficient.
And if the data centers pay their share, then this is that Goldilocks case where the electric bills get cheaper.
Okay, but you know that pay their share bit?
Greg says that's another balancing act.
You can get a good deal or you can get a bad deal, right?
Utilities and regulators often set a separate rate for large customers like data centers.
If the centers end up paying too little, then the rest of the utility's customers will be stuck paying more for those new power plants.
And so again, it comes to getting it right.
It comes to the regulators getting it right.
It comes to the utilities getting it right.
And kind of meeting that Goldilocks that we talked about, to build out the appropriate amount of capital while at the same time protecting ratepayers in order to have access to affordable and reliable energy.
So how likely is it that utilities will make good deals for residents?
That's what I asked Ari Pesko.
He's the director of the Electricity Law Initiative at Harvard.
So unfortunately the incentives here for the utility and the data center are not great from the public's perspective.
He says the reason for that is because utilities are competing right now to attract data centers.
Because the utility industry makes money by building infrastructure.
That's where their profits come from.
And there is now no better reason to build infrastructure than data centers.
And a great way to beat out another state or utility for that data center business?
Offer a sweetheart deal where the utility isn't paying all that much and residents are essentially subsidizing the cost with higher electric bills.
He says the one mitigating factor here is all the public backlash to data centers that could put pressure on utilities as they set these rates.
But Ari doesn't expect that to go too far.
After all, these utilities are typically monopolies.
There's a reason that for 150 years in this country, we have regulated monopoly providers.
Well, OK.
But you're right.
We have those regulations.
We have public service commissions and other groups, don't we?
Does that not solve this issue?
So utility regulation is hard in part because the utility controls all the relevant information.
Yeah, the utilities have way more information about their own operations.
Not to mention way more resources than cash-strapped regulators for things like lawyers and analysts.
So it sounds like to you.
This sounds like less of a Goldilocks zone to try and hit that could benefit consumers than sounds more like to go biblical on you for a second, putting a camel through the eye of a needle.
You know, maybe the hole's a little bit bigger.
Maybe we could choose a smaller animal.
But look, I think we have some safeguards, some protections.
They don't always work.
So, yes, it's possible for electric rates to drop if everything lines up right.
But that doesn't sound like a bet Ari is planning on making.
If you liked this episode, send it to a friend who might like it too.
Word of mouth is how we grow, so spreading the word is supporting our journalism.
