Ohio’s Data Center Power Rule Is What Other States Won’t Say Out Loud

Buckeyes for the win! (Adobe)

Data centers are the new “economic development win” that shows up in a press release before anyone asks the obvious question.

Where’s the power coming from?

And who’s paying to build the system that makes it possible?

Most places are still treating this like normal load growth.

It isn’t.

A single hyperscale campus can pull city-scale electricity. Not “eventually.” Now. And they want it firm, not “we’ll see what the grid can do.”

Then homeowners look at their bills, hear about new transmission, new substations, new upgrades, and they ask the only reasonable question:

Why am I paying for someone else’s private business to plug in?

Ohio is one of the few states dealing with that question head-on.

Their approach is pretty simple: the user pays.

No quiet subsidizing through everyone else’s rate base. No pretending the grid is infinite. No letting speculative projects squat on capacity like it’s free real estate.

If you want massive power, you either bring your own generation or you pay like you mean it.

The part that actually changes behavior is how Ohio handles reserved capacity.

Developers love to reserve big blocks of electricity early. It makes a site easier to sell. It makes a project look “real.” It creates leverage.

But a lot of these projects slip. Some get phased. Some get scaled down. Some never happen.

Utilities still have to plan and build as if the full load is coming, because the reservation is treated like a commitment.

Those upgrades don’t come free. And if the load doesn’t show up, the costs don’t disappear.

They get spread across everyone else.

That’s the subsidy. It’s just dressed up as “system investment.”

Ohio’s answer is basically: if you reserve the seat, you pay for the seat.

Large new data center customers are being pushed into rate structures that require them to pay for a substantial minimum amount of what they reserve, even if they don’t use it.

That does two important things.

It makes speculation expensive.

And it protects the public from funding infrastructure for a customer they’re not.

Ohio also made a clearer lane for behind-the-meter generation.

Translation: if you’re going to be a massive load, you can pair that load with dedicated generation behind your meter so you’re not leaning on the public grid as your default backstop.

This takes pressure off interconnection queues and reduces the need for “upgrades” that are really just custom-built service for one end user.

It also forces honesty.

Reliability costs money. If you want high uptime, you can build it. But you don’t get to socialize the bill while keeping the upside private.

Now, there’s an obvious pushback here: won’t this scare away investment?

Maybe the kind of investment that only works when somebody else eats the risk.

If a project pencils only when homeowners subsidize the backbone infrastructure, that’s not a “win.” That’s a transfer.

Ohio is basically saying: come build, but own the consequences of your demand.

That’s not anti-business. That’s just adult.

Here’s where it gets interesting, though, and where I think people are underthinking the next chapter.

Everyone is assuming data centers will keep demanding more and more power forever.

But what happens when they don’t?

Chips are getting more efficient. Workloads are getting optimized. Data centers are getting denser. Cooling is improving. Power management is improving. The industry has real incentives to do that because electricity is now one of the biggest constraints.

So imagine a world where a campus reserved a huge chunk of capacity, built on-site generation to secure it, and then five or ten years later the load profile changes.

Maybe they’re using less power per unit of compute.

Maybe they’ve shifted workloads.

Maybe the campus footprint grows, but total draw doesn’t rise the way the early forecasts claimed.

Now you’ve got a weird situation: a private company sitting on generation and capacity that isn’t fully utilized.

Unused power doesn’t just sit there politely. People start asking what you can do with it.

Can they sell it?

Can they wheel it back onto the grid?

Do we end up with a future where the biggest data center operators become power suppliers, not just customers?

Because once you normalize behind-the-meter generation at scale, you’re not that far from letting private load owners become quasi-utilities.

And then you get the question that sounds like a joke until it isn’t:

Are we going to be buying our electricity from Amazon one day?

From Microsoft?

From Google?

That might sound far-fetched, but the mechanics aren’t crazy. If you have generation, you have fuel contracts, you have interconnection, you have infrastructure, and you have a financial motive to monetize excess capacity.

There are regulatory walls today. There are market rules. There are public utility structures for a reason.

But rules move when the physical reality forces them to move.

If the grid is constrained and private players are the ones building generation fast, pressure builds to let that power flow where it’s needed.

And once that door cracks open, it doesn’t close easily.

So Ohio’s model solves one problem cleanly: stop ratepayers from subsidizing private megawatt demand.

But it also sets up a longer-term question nobody’s really ready to answer:

What happens when big tech owns both the load and the supply?

If you’re a regulator, you shouldn’t just be asking “who pays for upgrades?”

You should be asking “what kind of market are we building?”

If you’re a community, don’t just ask “will this raise my rates?”

Ask “who controls the power system ten years from now if private generation becomes normal?”

And if you’re another state watching Ohio, the real lesson is this:

Stop pretending capacity is free. Stop letting reservations act like commitments without consequences. Make the biggest users carry the real cost of what they’re asking for.

Ohio is doing that.

The next question is whether we’re comfortable with where it leads if those same users start owning the supply side too.

Because the day you hear someone casually say, “We’ll just buy power from the data center operator,” you’ll realize the grid debate has changed permanently.