ASHFALL INSTITUTE | SUBDUCTION ZONE
Building Bridges — Episode One
THE BARGAIN
P. A. Moore
Ashfall Institute | Subduction Zone
Written under the collaboration disclosed in Time Is a Wheel and We Are the Ground. Concept, argument and judgment: P. A. Moore. Research and composition: Claude Opus 5.
Ninety-nine percent.
That’s the share of political advertising spending in this election cycle that runs against data centers. Candidates in at least twenty-one races across eighteen states have put them on television, and the spending is split almost evenly between the two parties.
Almost nobody in the United States is paying to defend them. Not one side. Neither side.
Gallup puts it at seventy-one percent — that’s how many Americans say they don’t want an AI data center near where they live. More than three hundred bills have been filed in state legislatures in a matter of weeks. Fifty-four local moratoriums have already passed. Oklahoma has one running until November of 2029.
I want to start there, because I think a lot of people in this industry are about to make a serious mistake about what that number means.
They’re going to call it fear of technology.
It isn’t.
Let me tell you what people are actually angry about, and I want to do it properly, because I don’t think you can answer an argument you’ve refused to hear.
On the electrical grid that serves Ohio and twelve other states, demand from data centers pushed capacity costs up by nine point three billion dollars in a single year. That’s a hundred and seventy-four percent increase. It isn’t a talking point. It’s a number that arrives in the mail with your name on it.
The jobs don’t cover it. Good Jobs First — and they’ve been doing this work for twenty-five years — finds that the taxpayer cost routinely runs over a million dollars for every permanent job created. One million dollars of public money. Per job.
And the tax breaks are extraordinary. My own state, Georgia, is forgoing about two and a half billion dollars a year on data center incentives. Virginia is at one point six billion. Texas around a billion.
Then there’s the part that turns a bad deal into an insult. Most of these arrangements — the community benefit agreements, the ones that decide what a town actually gets — are negotiated behind closed doors and sealed under non-disclosure agreements. The people whose electricity bill pays for it are not permitted to read the contract signed in their name.
Add the water. Add the land. Add the noise, and what it does to the value of the house you were planning to leave to your children.
Now look at that seventy-one percent again.
That is not technophobia. That is a correct reading of a bad contract.
So here’s what I want to say to the industry, and I’d like to say it plainly, because I think the response so far has been a category error.
You have answered a structural problem with a communications strategy.
You have explained. You have run open days and published fact sheets about water recycling. You have pointed out, accurately, that the electricity would have to come from somewhere regardless.
And you have lost fifty-four moratoriums doing it.
The reason is that nobody is confused. They understand the offer. The offer is bad.
Here is the offer, stated honestly. A community is asked to accept a windowless building the size of forty football fields. It will consume the electrical output of a small city. It will pay little or no property tax for twenty or thirty years — an abatement the residents themselves subsidize. It will raise their power bills. It will employ perhaps forty people permanently, most of whom will be brought in. And they will not be allowed to see the terms.
Of course they said no. I’d have said no.
But here’s where I part company with almost everyone shouting about this right now.
None of that is a fact about artificial intelligence.
Every single item on that list is a fact about tax structure and rate design. About who a legislature decides should carry a cost. Those are written by people, in buildings, in daylight, and they can be written differently.
The machines are not the reason your bill went up. The rate structure is the reason your bill went up — the decision to socialize the cost of new capacity across every residential customer instead of charging it to the party that caused it.
The machines are not the reason your county is short two and a half billion dollars. The abatement is the reason.
The machines did not sign the non-disclosure agreement.
We are having a fight about computers when we should be having a fight about contracts, and I think a great many people find that substitution convenient.
Now. I don’t want to stand here and tell you the alternative is imaginary, because it isn’t. It’s already law in one state.
In May, New Jersey passed S731 and A796. What it does is technical and it sounds dull, and it is the most important thing that has happened in this fight.
It creates a separate ratepayer class for data centers.
That’s it. That’s the whole mechanism. It means a data center pays for its own electricity and for the grid infrastructure its arrival requires — instead of spreading the cost across every household on the system.
The principle has a name. Engineers and regulators call it cost causation. The party that causes the cost carries the cost.
California passed its own version at the end of August. There’s a voluntary federal pledge, announced in March and expanded in July, that asks companies to do this on their own — and Brookings has said the obvious thing about it, which is that a voluntary pledge needs enforcement to mean anything.
But New Jersey didn’t ask. New Jersey wrote it down.
And notice what that does to the seventy-one percent. It removes the electricity bill from the argument entirely. Not by explaining it. By fixing it.
I want to name the hard part, because otherwise this is just a wish.
Why does any state give a thirty-year tax abatement to a trillion-dollar company?
Not because the legislature is corrupt. Because the state next door is offering one.
That’s the trap. If Georgia ends its incentives tomorrow and Alabama doesn’t, Georgia does not collect two and a half billion dollars. Georgia loses the project and collects nothing. Every governor knows this. It’s a prisoner’s dilemma between fifty legislatures, and every one of them is playing it correctly from where they sit.
Which means any proposal that asks a single state to act alone is asking it to lose.
The answer has to be coordinated. An interstate compact. A federal floor beneath which abatements can’t go. A condition attached to federal energy or permitting support. We have done this before — the OECD global minimum tax exists precisely because thirty-odd countries recognized they were bidding each other down to nothing.
That’s a hard political lift. I’m not going to pretend otherwise.
But it is a solvable problem, and it is a completely different problem from the one we’re arguing about.
Here is the part I actually want to talk about, and it’s the reason for this series.
Everything I’ve described so far is defensive. Stop the bill going up. Stop the abatement. Open the contract. All necessary. None of it inspiring.
Nobody has drawn a picture of the good version.
That’s what the ninety-nine percent is measuring. It isn’t that people examined the optimistic case and rejected it. It’s that there is no optimistic case in circulation. There is a dystopia on television every night, funded by both parties, and there is nothing standing opposite it.
So let me put something opposite it.
In 1933 this country put three million young men to work. They planted three and a half billion trees. They built eight hundred parks, and trails, and bridges, and fire towers that are still standing. They were paid thirty dollars a month and twenty-five of it went home to their families.
At the same time, and for far less money, we hired writers. Federal One sent them out with notebooks, and among the things they came back with were two thousand three hundred interviews with people who had been born into slavery.
Nobody could have justified that on a spreadsheet. There was no return on investment. It was the last possible decade in which those interviews could be conducted at all, and somebody in a government office decided the country should have them.
We have them.
Here is what I want you to notice about both of those programs. What they built stayed. The trees stayed. The trails stayed. The narratives stayed. A hundred years later you can walk into one and read the other.
Now compare that to the current bargain.
A data center leaves a building and a bill.
That’s not an argument against data centers. We need them — I’ll say that plainly, and I’ll say more about why in the next episode. It’s an argument that we are asking for nothing, and getting nothing, and then being surprised that people are angry.
So that’s the question this series is about, and I’ll be at it for a while.
Not should we build this. We’re going to build it. That argument is over and it was over before it started.
The question is: what do we get?
And I mean that concretely. Not jobs numbers that don’t survive contact with an auditor. What stays. What’s here in fifty years that wasn’t here before. What the great-grandchildren of the people currently signing petitions will be able to walk into.
We have done deliberate reconstruction in this country before, at enormous scale, and we did it well enough that the results are still load-bearing.
We are about to spend more than that. Considerably more.
And at the moment, the entire plan is a windowless building and a non-disclosure agreement.
We can do better than that. I think we can do a great deal better than that, and over the next several episodes I’m going to try to describe exactly how — the architecture, the energy, the education, the work, and what it might mean to build something with a partner instead of installing something on top of a town.
Ninety-nine percent to nothing is not a debate.
It’s a forfeit.
Let’s stop forfeiting.
Next: why we actually need the buildings — and what a data center could give a county that a factory never could.
Building Bridges is a fourteen-part series. Next: Episode Two — Why We Need the Buildings.
P. A. Moore is the pen name of Pamela King, philosopher and artist. Available through the Ashfall Institute.