Definition

The decision a community makes about a data center is split across two venues that decide different things on different timelines.

The local body (city council, county commission, redevelopment commission) decides whether the project gets built — through zoning, tax abatement, water and sewer service, and the development agreement. The state utility commission decides who pays for the grid the project needs — through the rate case, often a year or more later.

The community’s leverage is concentrated in the first venue and nearly absent in the second. The development agreement is the one document where a locality can attach enforceable cost protections, because approval is conditional and the developer needs it. Once the local body votes yes, the cost-allocation question moves to a venue where the town has no standing beyond petitioning to intervene, and where it is one voice in a proceeding it does not run.

The window is the gap between the local approval and the rate case. Cost protections attached before the vote can be made binding; the same protections sought after the vote depend on a regulator the town does not control. The mechanism generalizes beyond formal “development agreements” to any discretionary local approval (a rezoning, a tax-abatement vote) where the build decision precedes the cost decision and lives in a different room.

Why It Matters for the Newsletter

This is the procedural reason data-center cost fights feel rigged to the people paying for them. The venue that decides their bill is the state rate case, which they cannot see and do not attend. The venue they can influence — the local council meeting — decides the build, not the bill. By the time the cost lands on the ratepayer, the leverage that could have blocked it is already spent.

For a TCN reader, this reframes AI Cost Incidence from a pricing question into a timing-and-venue question. “Who pays for AI’s grid buildout” is settled less by whether the rules are fair than by when and where the rules get written, and whether anyone with leverage was in the room. It also explains why plain-English disclosure at the local level is load-bearing rather than a courtesy: it is the only thing that lets a council use its leverage while it still holds it. A disclosure that arrives after the vote informs residents about a decision they can no longer shape — the same timing failure that defines Toothless Transparency Laws, in a different register.

The newsletter angle: a piece that follows a single local approval through to the rate case it triggers, showing the town go from holding the one card the developer wants to becoming just another name on the rate roll.

Evidence & Examples

Tensions & Counterarguments

  • The window is not a hard zero after the vote. Localities can sometimes petition to intervene in rate cases, and state AGs (Dana Nessel in Michigan) fight cost-shifting at the commission after the fact. Post-approval leverage is much weaker, not nonexistent. Don’t overclaim “the town has no recourse.”
  • Inter-locality competition bounds the leverage. A town that demands too much in the development agreement can push the project to a friendlier jurisdiction. The leverage is real but capped by the developer’s option to build elsewhere — the same dynamic that lets hyperscalers extract tax abatements.
  • By-right builds skip the window entirely. Where a data center can be built on existing industrial zoning without a discretionary approval, there is no leverage window to use. The model assumes a discretionary local gate.
  • Speed to Power bypasses both venues. Behind-the-meter / off-grid builds skip the interconnection queue and the rate case, which removes even the downstream check. There the local approval may be the only venue that ever scrutinizes the project at all.

Key Sources