Ten campuses, one county: opposition saturation in Lackawanna

The denial that stopped a $280 million data-center campus in Newton Township, Pennsylvania is one sentence long. On February 19, 2026, Zoning Officer D. Scot Haan reviewed Popca LLC's application for two buildings, 145,200 square feet each, 35 feet tall, on roughly 90 acres of the East Newton Ira Tripp Tract, and wrote that the proposed use is "not an allowed use within Newton Twp."

No hearing. No traffic study. No consultant fees on the other side.

The Rural Resource zoning district does not enumerate data centers, so the answer was no, and the entire fight now runs through the appeal Popca filed on March 23, 2026. The hearing is scheduled for April 27, 2026, at 7 p.m. at the Newton Township Municipal Building. Our record closes with that hearing on the calendar, not held, and a scheduled hearing is a date, not an outcome.

What makes Newton worth a research note is not the one-sentence denial. It is the number ten.

What does saturation do to a single application?

Per WVIA's April 2026 reporting captured in the Lackawanna record, ten data-center campuses are proposed across Lackawanna County: six in Archbald, and one each in Clifton, Covington, Dickson City, Newton, and Ransom townships. Ten simultaneous proposals in one county changes the physics of every individual hearing. Opposition bandwidth turns out to be a regional pool rather than a township resource, and the record shows it working exactly that way: residents who organized against the Ransom Township proposal testified in Newton's hearings. Every applicant in the county is now paying for the other nine, because each hearing has stopped being a land-use question and become a referendum on the countywide pattern.

The sequence bears that out. In January 2026, Ransom Township supervisors rejected the zoning amendment that would have enabled Scranton Materials LLC's six-building campus at 819 Newton Road. Attorney Michael Mey appealed that denial to the Lackawanna County Court of Common Pleas in March 2026. Weeks after the Ransom board said no, Newton's zoning officer issued his one-sentence denial next door. Two adjacent townships took different procedural weapons to the same destination.

The Newton application itself deserves a moment, because it was not a speculative sketch. Popca LLC, an Archbald entity with Mark Gawron as the prior property owner contact, had acquired the tract in 2025. The filing described two buildings of 145,200 square feet apiece, roughly three acres of roof each, at an estimated $280 million. Somebody spent real money getting to a one-sentence no.

We scored the Newton parcel 28 out of 100, and the two load-bearing deductions were visible before any filing: Rural Resource zoning that never mentioned the use, and a county already saturated with proposals it had begun refusing. A textual denial is the most preventable kind there is. The ordinance was public, the parcel's district was public, and the absence of a data-center use in the Rural Resource enumeration could have been confirmed for the cost of an afternoon before the land changed hands.

Why does the type of denial matter as much as the denial?

Because the two denials, which look identical in a headline, produce opposite remedy paths, and the remedy path is where the money goes.

Newton's denial is administrative. A zoning officer read the ordinance text and found no data-center use enumerated in the Rural Resource district. That kind of denial gets appealed to the zoning hearing board, and it can in principle be cured legislatively, by a text amendment. Watch what the legislative body does with that opening. Newton Township's supervisors had initiated no text amendment as of April 2026, and the silence is the signal: when a board wants a land use, it amends the code. When it leaves its zoning officer's textual reading standing, the reading is the policy.

Ransom's denial is legislative. Supervisors voted down the amendment itself, and the only recourse from a legislative zoning decision is the county court, where Pennsylvania judges give township supervisors heavy deference absent a clear abuse of discretion. That appeal is now pending at Common Pleas, and appeals of this kind are multi-year processes. A capital schedule built around GPU delivery windows cannot absorb a multi-year detour through a deferential court.

And the appellate math is unforgiving even when it goes well. Suppose Popca wins on April 27 and the hearing board overturns the zoning officer's reading. Per the record's own hedge, a favorable ruling would itself be subject to further appeal by aggrieved neighbors, and in a county where the neighbors are already organized across township lines, the aggrieved-neighbor appeal is close to a certainty. The applicant's best case is a win that buys another round.

A one-sentence administrative denial and a board-room legislative denial cost the same to receive and wildly different amounts to fight.

There is a rhyme here with the Indiana pattern we documented in the moratorium contagion file, where four rural counties blocked data centers in ten months, each borrowing nerve from the last. Lackawanna is the compressed version: the contagion does not need to jump county lines, because all ten proposals landed inside one county's borders at once. Indiana's opposition learned from its neighbors. Lackawanna's opposition car-pools.

Where does the capital go instead?

The record itself points at the exit. Lackawanna County has industrial-zoned land near the existing Lackawanna Energy Center and along established industrial corridors, where data-center use is either enumerated or compatible with what the code already says. The denied parcels were never the only inventory. They were the inventory whose zoning risk got price-tagged at exactly one sentence.

That is the discipline saturation forces on site selection. In an empty county, a marginal parcel with hopeful zoning might be worth the entitlement gamble, because the applicant controls its own narrative. In a saturated county, the marginal parcel inherits the political weather of every other application in the pipeline, and the weather in Lackawanna as of spring 2026 was two denials on appeal and eight proposals still in the queue, in a county whose boards have watched their neighbors say no. Screening a data-center site in a county like that starts with counting the pipeline, not walking the land.

Count the campuses before you count the acres.

My read is that county-level saturation is now a first-order screening variable, on par with the zoning text itself, and the market has not caught up to that. I expect the next Lackawanna denials to cite the countywide pattern more or less openly, and I expect at least one of the ten applicants to withdraw before its hearing rather than spend into saturated opposition. If the April 27 Newton appeal instead produces a reversal and the county pipeline starts clearing, I got the direction wrong, and the correction will be worth writing. Watch the number ten. Whichever way it moves, it moves the odds for the other nine.

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.