I sat in a meeting recently where a group of mayors were working through what they could do about projects proposed in the unincorporated county around them. Someone raised water service. Someone else raised fire code. Someone else raised road standards.
All three are real. All three are also things a well-capitalized applicant can pay for, on its own schedule, without asking anyone’s permission.
That’s not a reason to give up. It’s a reason to get precise, because there is a category difference between a requirement a developer must satisfy and a decision a public body gets to make — and communities routinely spend their political capital on the first while quietly forfeiting the second.
The distinction that organizes everything
Ministerial approvals are the ones an official must grant if the applicant meets the standard. Building permits. Driveway permits. Utility connections where a duty to serve exists. The official’s job is to check compliance, not to weigh the merits.
Discretionary approvals are the ones a public body gets to decide. Rezonings. Conditional use permits. Comprehensive plan amendments. Here the body weighs, and conditions attach.
Ministerial approvals are not leverage. Withholding one to extract a concession isn’t hard bargaining — it’s an abuse of authority, and it’s how a county ends up defending a hundred-million-dollar lawsuit instead of negotiating a benefits package.
So the very first thing to establish is which of your gates are actually gates.
The five gates, in order of value
The comprehensive plan. Earliest and cheapest. If the plan doesn’t designate the site for intensive industrial use, a rezoning is inconsistent on its face — and that inconsistency is the strongest legal footing a denial can stand on.
Rezoning. Maximum leverage, because it’s legislative and discretionary. A denial is hard to overturn absent arbitrariness.
The conditional use permit. Nearly as strong, and better in one specific way: conditions attach and the permit remains revocable across the facility’s operating life. Every requirement you want enforceable in year fifteen belongs here.
Utility service and access. The underrated one. Water and sewer service outside corporate limits is typically a contract, and contracts carry conditions. So do road access, driveway permits, easements and crossings. Note the qualifier, though — where a certificated provider has a statutory duty to serve, this is ministerial and you should say so rather than pretend otherwise.
Incentive approval. The most visible gate and often the least powerful, because its leverage is entirely conditional on the developer needing the money. Incentives run roughly two percent of hyperscale construction cost. A project with a signed tenant and a queue position frequently does not need your abatement, and if you’ve built your entire strategy on it, you’ll find that out at the worst possible moment.
Now the uncomfortable part
If a developer is willing to pay one hundred percent of the cost of water infrastructure, fire apparatus, road improvements and emergency response — and to contract for all of it on its own timeline — most of what a community thought was leverage evaporates.
Not all of it. But most.
This is worth saying plainly to a community rather than letting them discover it in month six. A lever that can be bought out is not leverage. It’s a price. And once it’s been paid, the conversation you were planning to have is over.
The productive response isn’t to pretend otherwise. It’s to know which items are prices and negotiate them as prices — briskly, for as much as they’re worth — while spending your real capital on the things money can’t resolve.
What actually can’t be bought
Four things, and they’re more valuable than they look.
Time. Nearly every applicant is on a schedule set by chip deliveries, tenant commitments and interconnection dates. Nothing a community does costs a developer more than uncertainty about timing. This is not a reason to stall — stalling in bad faith is both wrong and legally hazardous — but it is a reason to understand that a clear, fast, well-run process is genuinely valuable to an applicant, and that value is tradeable.
The record. What gets put in writing, published, and made findable is permanent, free, and outside anyone’s ability to purchase. A standing project page with every document on it costs a jurisdiction nothing and no developer’s cooperation.
Reputation. Not the community’s — the applicant’s. These companies are siting the next project somewhere, and the next one after that. How they behave here becomes the case study cited there. That is a real asset that a developer will spend real money to protect, and it is the actual engine behind most generous community agreements.
Capability. A jurisdiction that knows what to ask for gets more than one that doesn’t, on identical legal authority. The twelve project questions, the conversion of adjectives into clauses, the ordinance adopted before an application arrives — none of that requires you to hold a veto.
The version I give to officials
If you have discretionary approvals, you have real authority, and you should use it deliberately and lawfully — with conditions in the permit, not just promises in a contract.
If you don’t, you still have four things: a process that determines how fast this goes, a public record that determines how it’s remembered, an applicant that cares how it looks doing this, and the capacity to know exactly what to ask for.
That’s not nothing. It’s frequently what produces the better agreement, because the applicant that has to be forced gives you the minimum and resents it, while the applicant that wants to be seen doing this well will go further than your leverage could ever have compelled.
And the version I give to developers
The same slide, read from the other side.
If your entire community strategy assumes you can buy your way past every local constraint, you are optimizing for the wrong risk. You probably can. And then you’ll operate for twenty years in a county whose next ordinance was written specifically about you, in front of a successor board that inherited the grievance, with a complaint file that starts on day one.
The gates aren’t the real constraint. The record is.