Data Centers, Farmland, and the Debate We're Not Having

Andy Masley got something important right: most of the land use concerns swirling around data centers aren’t really about land.

America is not running out of it. Agriculture is not beyond scrutiny. And most of the people invoking farmland at planning commission meetings are not doing analysis; they’re misdiagnosing a governance problem as a land problem.

What he doesn’t name is why that keeps happening.

His piece matters — and I want to say that clearly — because it pushes the conversation away from emotional symbolism and toward the structural questions underneath. Even where I land differently than he does, he’s right that the debate is reacting to surface-level imagery instead of the actual tensions forming beneath it. That’s a conversation communities, planners, utilities, and county commissioners are going to have to get much more serious about. Fast.

But here’s where I want to pick it up and go further.

I’ve spent over two decades inside planning departments, county courthouses, and state regulatory processes. Communities that range from small to large incorporated towns to rural counties with no planning staff at all. I’ve sat across tables from commissioners trying to figure out what a 500-acre industrial site means for their road maintenance budget and their volunteer fire department.

What I can tell you is this: the real scarcity in development is almost never the land itself.

It’s the combination. Transmission access. Substation capacity. Fiber proximity. Adequate water. Industrial zoning. And, maybe most critically, local institutions that actually understand what they’re approving.

A thousand acres in the abstract doesn’t mean much. A thousand acres next to a transmission corridor, with a utility willing to serve and a commission that can evaluate what they’re looking at? That’s something else entirely.

Call it infrastructure geometry. That’s where the real development conversation lives. It’s also where most public debate never arrives.

So when opponents say a data center is “taking” farmland, they’re usually describing something much simpler: a landowner who sold voluntarily, in a location that happened to meet a very specific infrastructure checklist that most land in this country simply doesn’t. That’s not a land crisis. That’s an infrastructure geometry problem.

And supporters aren’t getting it right either when they reduce the whole thing to tax revenue per acre. Every large land use decision creates pressures you don’t see until they show up — utility load shifting away from neighboring communities, transmission expansion that bypasses rural corridors, emergency response demands that outpace volunteer capacity, changes to who a place thinks it is.

Rural jurisdictions especially are operating with limited planning staff, fragmented utility coordination, and no long-range infrastructure modeling. In that environment, a hyperscale data center doesn’t arrive like economic development.

It arrives like weather.

That’s not irrational opposition. That’s governance lag. And governance lag is and has been one of the defining development problems of this decade.

I watched it happen recently. A 227,000 square foot warehouse (vacant about 80% of its life) sits in a small Oklahoma city where the municipality is the utility provider for everything: water, sewer, electric, trash. On the property line: a substation. Someone came forward wanting 10,000 square feet of it for a tier 1 data center. Less than five percent of the building. The city would have captured utility revenue directly. The data center footprint would have left 90 percent of the building available for a tenant who could actually bring jobs, with infrastructure costs largely offset by what the data center was already generating.

The city placed a moratorium until the end of the year.

Not bad faith. The data wasn’t there, and the applicant didn’t make the case well enough for a council that had no framework to evaluate it. So the opportunity sits. The substation sits. The building sits.

That’s governance lag, not malice or ignorance. A system that hasn’t yet learned to recognize what it’s looking at.

This isn’t a new pattern. Small towns welcomed large-scale retail consolidation for the same reasons: the immediate economics made sense. Lower prices, sales tax revenue, new construction. But over time, local economies became dependent on decision-makers with no long-term ties to those communities. Decisions about the economic survival of entire towns got made in distant boardrooms by people optimizing for national efficiency, not local resilience.

The question isn’t whether data centers are good or bad. The question is whether communities understand the structural position they’re entering when they agree to host infrastructure whose economic gravity and decision-making sits somewhere else.

And this is where I want to go further than Masley.

The acreage question, the tax revenue question, even the infrastructure geometry question… these are all technical arguments about something that isn’t really technical at its core.

What we’re actually debating, underneath all the farmland rhetoric, is what kind of society we’re building.

Hyperscale data centers aren’t just real estate decisions. They’re decisions about where computing power concentrates, who controls it, how it’s governed, and which communities get to participate in the AI economy versus simply hosting its physical infrastructure while the value moves elsewhere.

A rural county that permits a hyperscale facility and receives assessed valuation has made a trade. Whether it’s a good trade depends on questions that almost never get asked at the planning commission: Who owns the intelligence running inside those buildings? Who captures the productivity gains? What happens to this community in ten years when the infrastructure it hosts is operated remotely, maintained by specialists who fly in, and controlled by people who’ve never set foot here?

Those aren’t anti-technology questions. They’re governance questions. They should be at the center of this debate.

They’re almost never there.

Instead we argue about farmland.

But that future isn’t fixed. Some communities have already figured out what it looks like to treat digital infrastructure as a platform for something bigger than tax revenue.

Chattanooga did it. The city used municipally controlled fiber through its public utility, EPB, to build one of the fastest broadband networks in the Western Hemisphere. The long-term value wasn’t internet speed — it became a foundation for smart grid modernization, startup ecosystems, research partnerships, workforce development, and a regional identity built around technology. A 2025 economic impact study put the community benefit at over $5.3 billion since 2011, on a roughly $280 million original investment. Chattanooga didn’t just host infrastructure. It used infrastructure to change its position.

San Antonio is navigating it right now. Council members are pushing for proactive frameworks on water consumption, power load, zoning buffers, and cost allocation — trying to get ahead of infrastructure demands before they outpace the institutions meant to manage them. That’s the right instinct.

And Loudoun County, the largest concentration of data centers in the world, shows both sides of what’s possible. Data centers are projected to generate nearly $1.3 billion in county tax revenue in fiscal year 2027, roughly 45 percent of total county revenue. Real property tax rates have dropped every year since 2016. But the tensions around energy demand, transmission expansion, and community identity are real too. In March 2025, the Board of Supervisors ended by-right zoning for data centers and required public hearings on new projects for the first time. That’s governance capacity catching up to infrastructure reality.

The land use concerns being raised are largely misidentified. I’d go further: they’re symbolically displaced. A way of expressing legitimate anxiety about rapid infrastructure change without the vocabulary to name what’s actually frightening.

What’s frightening isn’t the acreage.

It’s the possibility that communities end up hosting the infrastructure of the AI economy without shaping it. Absorbing the costs (the transmission corridors, the water demands, the governance strain, the political disruption) while ownership of the models, the intellectual property, the computational leverage, and the compounding productivity gains concentrates somewhere else entirely.

That’s the real land use question.

It was never about farms… written by a girl who grew up on one.

Sources

1. Community Networks, Chattanooga’s Municipal Fiber Network Has Delivered $5.3 Billion in Community Benefits (December 2025) — https://communitynetworks.org/content/chattanoogas-municipal-fiber-network-has-delivered-53-billion-community-benefits-new-study
2. GovTech, Fiber Broadband Seen as a Force Multiplier in These Cities (December 2025) — https://www.govtech.com/network/fiber-broadband-seen-as-a-force-multiplier-in-these-cities
3. San Antonio Report, Risk and Opportunity: San Antonio’s Data Centers and Water and Electricity Systems (March 2026) — https://sanantonioreport.org/risk-opportunity-san-antonio-water-electricity-data-centers/
4. Texas Public Radio, How Should San Antonio Regulate Data Center Growth? (March 2026) — https://www.tpr.org/podcast/the-source/2026-03-29/how-should-san-antonio-regulate-data-center-growth
5. City Journal, Loudoun County, Virginia: The Heart of the Data-Center Boom (May 2026) — https://www.city-journal.org/article/loudoun-county-virginia-data-centers-construction
6. Loudoun County Official Website, Data Centers in Loudoun County — https://www.loudoun.gov/6188/Data-Centers-in-Loudoun-County
7. MultiState, State Data Center Legislation Faces Local Zoning Battles (January 2026) — https://www.multistate.us/insider/2026/1/15/state-data-center-legislation-faces-local-zoning-battles
8. https://blog.andymasley.com/p/data-center-land-use-issues-are-fake?utm_source=substack&utm_campaign=post_embed&utm_medium=web