This week in Matthews, a glimpse into the potential future of North Carolina’s economy came in the form of a rezoning case.
Developers pitched a plan to turn a tract of wooded land near downtown into a massive data center. The idea makes sense: Data centers are the plumbing of modern life. Every map search, bank transfer and AI query runs through them — so the demand is booming.
But neighbors aren’t so sure. They packed the meeting room, not to sneer at technology, but to ask a simple question: Is this good for the town? Matthews wants homes and storefronts on this stretch of road, not an industrial power hub. People want straight answers on power, water and noise.
Matthews isn’t alone. Similar projects are underway across North Carolina, in big cities and little towns alike. All different towns, but the same dilemma: Are data centers good for us?
The third wave
To try to answer that question, I talked to Scott Millar, the economic development director of Catawba County and the man responsible for recruiting plenty of data centers. So many, in fact, that his region got dubbed the “Data Center Corridor.”
His answer was nuanced, because not all data centers are the same. We are now solidly in the third wave of data center development, and by far the most controversial.
The first wave was small and company‑owned data centers — think back‑office rooms that host data for Bed, Bath & Beyond.
The second wave was based on cloud computing, and run by names everyone knows: Microsoft, Apple, Amazon.
This third wave is different, and it’s two things at once: enormous AI megasites and a rush of developer‑driven projects trying to cash in on the boom. That second category demands more scrutiny, especially when it’s unclear who the end user will be.
When Microsoft came in to consider sites in Catawba County, Millar told me, the company presented specifics: how much light would show, how much noise there would be and how much water it would draw. People could get comfortable with the project. If you don’t know who’s moving in, you can’t credibly answer the neighborhood’s most basic questions.
“People need to analyze the pros and the cons for these facilities,” he added. “It’s every community’s right to decide.”
Put it in the right places
These decisions are coming all across the state. To name just a few examples: In Mooresville, a $30 billion proposal on Teresa Earnhardt’s land died under an avalanche of opposition. In Tarboro, a 50‑acre plan was rejected after residents raised environmental and quality‑of‑life concerns. But not all of them were defeated. Statesville just approved a five‑building campus on former farmland after weeks of packed meetings. In west Charlotte, the city signed off on a 156‑acre rezoning that could become the state’s largest data center.
Let’s also be honest about what these facilities are and are not. They’re not mills or auto plants. They won’t employ by the thousands, but they also won’t flood neighborhoods with traffic once construction ends. During the buildout, there’s going to be construction and mud, but once that’s done, the day‑to‑day impact is light. You might have some box trucks bringing in computer servers, but not a bunch of semis rolling in.
They also offer a small number of solid, specialized jobs, and they can be powerful engines for a town’s tax base if sited well. That could be exactly the lifeline former factory towns need: steady revenue to fund parks, public safety and schools long after the ribbon-cutting.
Maiden is the proof. All the Catawba County data center investment there helped pay for a new community center, a new town hall, new police and fire headquarters and a satellite fire station all while the town lowered its tax rate.
Say yes, maybe
So what should Matthews, and the rest of North Carolina, do now? Start with a clear standard and stick to it. Say yes to data centers, but say no when the fit is wrong.
Data centers aren’t auto plants. They bring huge capital investment and years of construction activity, but relatively few permanent jobs. That doesn’t make them bad; it means we should judge them on the right scale — long‑term tax base, infrastructure fit, and whether they strengthen or strain the communities around them.
To determine that, towns can insist on a named operator, hard numbers on power, water and noise and sites built to carry the load. Do that, and data centers become a tool to rebuild places hit hard by factory closings, not a burden on the ones that are already thriving.
Welcoming this industry does not require trading away peace and quiet. Done with care, data centers can underwrite local budgets and revive dormant sites without swallowing what makes a place home.
Projects that can name an operator and show their math on power, water and noise deserve a real hearing. Projects that can’t answer basic questions don’t belong next to small‑business corridors or pressed against subdivisions.
This site in Matthews doesn’t seem to fit the bill. But in many other towns around the state, data centers will make sense.
Data centers are coming. North Carolina should welcome them — on our terms.
Contributing columnist Andrew Dunn is the publisher of the Longleaf Politics newsletter, which offers thoughtful analysis of North Carolina politics and policy from a conservative perspective. He can be reached at andrew@longleafpol.com.
