The moratorium is contagious: how data-center pauses jump county lines

In February 2025, Marshall County became the first county in Indiana to pass a data-center moratorium. Here is the detail that should stop every site-selection team cold: there was no data center to stop. No pending application. No hyperscaler at the podium. Commissioner Stan Klotz explained the vote himself: "Basically, we didn't like what we saw going on around us in other areas, and we just weren't quite totally sure how to respond."

Read that again. The trigger was not a project. It was the neighbors.

That is the mechanism this post is about, because it is the mechanism most pro formas still price at zero. Moratoriums are not a local response to a local application. They are contagious. They spread county to county on fear of the fight next door, they escalate once adopted, and by the time your land team notices one on a county agenda, the political conditions that produced it have usually been visible in the public record for months.

Why do counties pass data-center moratoriums with no application pending?

Because a moratorium is the cheapest political product a county commission can buy. It costs nothing, it defers every hard question, and it reads to voters as prudence rather than obstruction. The Indiana county record lays the sequence out like a lab experiment.

Marshall County moved first, in February 2025, defensively. White County followed on October 20, 2025, by a 2-1 vote. Putnam County passed a one-year hold in November 2025, and bundled it: wind, solar, data centers, and small modular reactors, all under one pause. By November 26, regional press was already reporting the three counties as a pattern. Then Starke County's Planning Commission voted unanimously on December 4, 2025, in a meeting relocated to the Knox High School auditorium to fit the crowd, to recommend a one-year moratorium on any data-center proposal over 5,000 square feet. Final adoption went to the county commissioners on December 15.

Four counties in ten months. One actual proposal chased them? No. The contagion ran ahead of the applications.

Two accelerants make the spread faster than the underlying politics would manage alone. The first is advocacy infrastructure: Citizens Action Coalition, a multi-county organization, connects isolated rural commissioners into a statewide posture. Its organizer Bryce Gustafson put the doctrine plainly: "Once the barn door gets open on that stuff, it makes it easier for the horses to get through." That is a zero-tolerance frame, and it means the first approval anywhere is treated as a permanent market opening everywhere. The second accelerant is pattern coverage. Once the press writes the "wave" story (Georgia Public Broadcasting ran exactly that framing for Georgia's county ordinances), every commission that has not yet acted gets asked why not.

Do moratoriums stay temporary?

The honest answer from the record: they escalate more often than they expire quietly.

DeKalb County, Georgia is the cleanest escalation curve I have seen. The county had issued zoning certification letters confirming data centers were a permissible use on two light-industrial sites. Residents near Bouldercrest Road organized around generator noise, air quality, heat, and the concentration of industrial uses in predominantly Black neighborhoods, an environmental-justice framing that travels much further than standard land-use objection. On July 8, 2025, the commission passed a 100-day moratorium by a bare 4-3 vote. By December 16, 2025, the extension through June 23, 2026 passed unanimously. All three original no votes flipped. In five months, opposing the pause went from a defensible commission position to a politically unavailable one, and the county is now drafting a five-tier data-center ordinance covering size, spacing, noise, power, and water.

Starke County makes the escalation explicit rather than implied. Planning commission vice president Mark Allen, on the record, December 4: "I have a feeling this will be... this moratorium will go on and then we're going to take the time and put probably some type of stop if we can by statute in Indiana." A second commissioner, Howard Bailey, separately voiced support for a long-term or permanent ban. When the people running the pause tell you the pause is a placeholder for a ban, believe them.

And then there is the speed run. In Fayetteville, Georgia, Crow Holdings' entity brought a site plan for a parcel where data centers were permitted by right under Business Park zoning. The Planning and Zoning Commission denied it on January 27, 2026, with roughly 100 residents in the room. On March 5, 2026, the City Council adopted Ordinance 26-O-12, prohibiting new data centers in every zoning district in the city. Thirty-seven days from first hearing to citywide prohibition. The developer withdrew its appeal on March 18, one day before the council was scheduled to hear it, because there was nothing left to appeal into. By-right zoning did not slow the ordinance down at all; the use table is only as durable as the council's willingness to leave it alone.

What does saturation look like when the wave arrives all at once?

Lackawanna County, Pennsylvania is the stress test. Per April 2026 reporting, ten data-center campuses were proposed across the county simultaneously: six in Archbald, one each in Clifton, Covington, Dickson City, Newton, and Ransom townships.

Opposition bandwidth turned out to be a regional pool, not a township one. Residents who organized against the Ransom Township campus testified in Newton Township's hearings. Ransom's supervisors rejected the enabling zoning amendment in January 2026. Newton's zoning officer denied a $280 million, two-building proposal on February 19, 2026 with a single textual finding: a data center is "not an allowed use within Newton Twp." Both denials are on appeal, one at the county Court of Common Pleas, and Pennsylvania courts give township legislative decisions heavy deference. Every additional campus in the county raised the temperature on all the others. Ten simultaneous proposals did not spread the opposition thin. They trained it.

How should a site-selection team price this?

Three adjustments, all of them cheap compared to what they prevent.

First, screen the neighbors as hard as the parcel. Marshall County proved a moratorium needs no local trigger. If the adjacent county is fighting a data center, your county's commission is already drafting talking points. A jurisdiction read that stops at the county line is measuring the wrong radius.

Second, treat any active moratorium as a one-way door until proven otherwise. The DeKalb curve (split vote, then unanimous extension, then ordinance drafting) and the Starke pre-announcement both say the same thing: the pause is the beginning of a rule-writing process, and the rules get written by whoever shows up while you wait. If you plan to build in a paused county, the moratorium window is when the engagement work happens, not when the deal sleeps.

Third, stop treating by-right as a terminal answer. Fayetteville converted "permitted use" into "prohibited in every district" in 37 days. The real screening question is not "is the use permitted today?" It is "how fast can this jurisdiction's legislative body amend the code, and is there an organized base that would push it to?" One of those questions is in the use table. The other is in the public record RealClear reads before your budget moves.

The state-preemption caveat belongs on the table too. Indiana's HB 1333, which would have given localities a 1% share of the sales-tax exemptions, passed the House 54-45 and stalled in the Senate, and no statewide framework emerged. Preemption could, in principle, flatten the county map. It has not yet, anywhere in this record.

So here is my call, on the record. I expect Starke County's moratorium to convert into a permanent or near-permanent restriction rather than lapse when its year runs out in late 2026; two of its planning commissioners have already said so out loud. And I expect at least one more state to produce a four-county-in-a-year cluster like Indiana's before the end of 2027, because nothing about the contagion mechanics is Indiana-specific: the politics are free, and the advocacy infrastructure travels. What would prove me wrong is a clean statewide preemption bill actually passing somewhere. Watch the state legislatures. Until one of them moves, the county map is the market.

This analysis is a source-cited research summary drawn from public records, not legal advice. It can contain errors and should be verified independently before any investment decision.