Why This Is Not a Passing Phase

It is tempting to read the current wave of data-center opposition as a temporary spasm, a burst of unfamiliarity that will fade as communities get used to these facilities and as developers get better at outreach. Under that story, the empty hearing room eventually returns, and consent goes back to being a formality. I want to argue against that story, because I think it is wrong, and believing it will get you hurt. The evidence points the other way: the consent constraint is tightening, and the forces driving it are structural rather than transient.

Consider the trend in moratoriums. A few years ago, a data-center moratorium was an exotic event. Now the debate appears on agendas across many jurisdictions, and each one that passes does two things. It stops projects in that jurisdiction, and it serves as a template for the next jurisdiction, because local officials borrow ordinances from one another constantly. A moratorium is not just a local pause; it is a piece of transferable technology in the opposition's toolkit, and transferable technology spreads. The base rate of a county being willing to hit pause on the whole category is rising, and a rising base rate is the opposite of a passing phase.

Consider the professionalization of opposition. Early resistance was disorganized and procedurally naive, which made it easy to outmaneuver. That is no longer the situation. Residents now share tactics, ordinances, and expert contacts across jurisdictional lines. They arrive at hearings fluent in the comprehensive plan and the noise ordinance. They know how to request the continuance and the additional study that convert a likely approval into a stalled one. This kind of capability does not decay once built; it accumulates and transfers. Each fight trains the participants and produces materials the next fight reuses. An opposition that learns is an opposition that gets harder to beat over time, not easier.

Consider, most fundamentally, the costs themselves. The concerns driving opposition, electricity rates, water, noise, air, land-use character, fairness, are not misunderstandings that education dissolves. They are, in the main, real costs or real risks that scale with the size of the projects. And the projects are getting larger, not smaller. As individual campuses grow toward the power appetite of a small city, the externalities grow with them, and so does the number of residents who notice. You cannot outreach your way out of a cost that is genuinely increasing. The structural driver here is that the thing communities are worried about is getting bigger, which means the worry is rational and durable rather than a temporary misperception.

There is a political economy point layered on top of the physical one. Local land-use hearings are one of the few venues where ordinary residents hold direct leverage over very large companies, and people do not voluntarily surrender leverage once they discover they have it. The asymmetry that usually favors capital inverts in that room, and the residents have noticed the inversion. Expecting them to forget it is expecting a rational party to abandon a source of power for no reason. That is not how leverage works. If anything, the discovery that the hearing room is powerful will draw more participation, not less.

Now, I want to be careful not to overclaim, because an empiricist who overclaims is just a pessimist with footnotes. I am not saying data centers will stop being built. They will keep being built, because the demand behind them is enormous and real. What I am saying is narrower and more useful: the constraint on where and how they get built has shifted durably toward community consent, and that constraint will keep tightening for as long as the projects keep growing and the opposition keeps learning. The equilibrium is not no data centers. It is data centers, but only where consent can be genuinely earned, and increasingly not where it cannot. That is a real and permanent change in the siting landscape, and it rewards the developers who internalize it early.

The practical implication follows directly. If consent were a passing phase, the rational response would be to wait it out or to power through with messaging. Because it is not a passing phase, the rational response is to build the discipline of earning consent into the permanent architecture of how you site projects, the same way you built in the disciplines of power procurement and fiber diligence. You do not treat the grid as a temporary inconvenience to be messaged past. You will not, if you are wise, treat consent that way either. In the final lesson I want to translate this into a posture and a practice you can carry out of the course.