AI Data Centers and Solar Farms Are Eating the Midwest's Best Land

For decades, “the land grab” in farm country meant foreign sovereign wealth funds, private equity, and corporate agriculture buying up the family farm. Now there’s a new category of buyer — and they’re not here for the corn.

The June 20 edition of Farm Progress’s “This Week in Agribusiness” flagged data centers and solar farms as featured topics alongside commodity programs and grain bin safety — a signal that in farm country, energy infrastructure’s appetite for land has become standard agenda-item territory, right alongside crop prices and equipment costs.

It shouldn’t surprise anyone who’s watched the map. Flat, lightly developed, power-grid-adjacent agricultural land in the Midwest checks every box on a data center or utility-scale solar developer’s site selection list: low cost per acre, proximity to high-voltage transmission lines, access to water for cooling, and far fewer zoning obstacles than developed suburbs. The same geography that makes great corn country makes great infrastructure country. And right now, infrastructure capital has more money than farmers do.

The Data Center Land Race

The artificial intelligence buildout has driven an unprecedented wave of data center construction across the United States. Tech companies began announcing major AI infrastructure investments in 2023, and the Midwest absorbed a disproportionate share of the new campuses — particularly Iowa, which now hosts large data center footprints from Microsoft, Google, and Meta alongside its traditional farm-economy anchor industries.

Data centers don’t look like heavy industry. From the road, a hyperscale campus looks like a cluster of large warehouse buildings inside a security fence. But their land and power footprints are enormous. A single hyperscale complex can require hundreds of acres and draw enough electricity to power a small city. When those campuses displace agricultural ground, that land typically doesn’t come back into crop production — the infrastructure investment is designed for a 20-to-30-year horizon.

Reuters and AP have tracked the scale of this buildout extensively. What those reports tend to underemphasize is where the land comes from. Much of it comes from the rural agricultural belt, sold or leased from farming families facing exactly the financial pressures — debt, input costs, uncertain markets — that the rest of this desk covers.

The Solar Overlap

The solar picture is different in character but similar in consequence. The U.S. Energy Information Administration tracks utility-scale solar capacity by state, and growth in the agricultural Midwest has accelerated sharply. Unlike rooftop solar — installed on existing structures, no land conversion — utility-scale installations typically require large cleared fields. A 200-megawatt installation can cover thousands of acres. Multiple installations in the same region can meaningfully remove land from agricultural production.

Solar developers have been offering farmland lease rates that can exceed what row crops generate per acre in years of suppressed commodity prices. For a farm family already carrying heavy debt or facing succession pressures, a solar lease or outright sale to an energy developer can look like the most rational exit available. The Chapter 12 bankruptcy data this desk has been tracking makes clear how many operations are facing exactly that financial crossroads.

The Permanence Problem

What makes this pattern particularly consequential is permanence. When land is paved for a data center or fenced off under a 30-year solar lease, it exits food production for at minimum a generation. Leased land reverts eventually, but the infrastructure left behind — grading, foundations, fencing, power connections — complicates return to row-crop use. And if commodity markets recover, or if farming technology makes a previously marginal operation viable again, that land isn’t available.

USDA’s Economic Research Service has tracked farmland value and land-use shifts for decades. The pattern of agricultural land converting to development at the urban fringe — well-documented since the postwar era — is now extending deeper into rural America wherever energy infrastructure demand reaches. The drivers have changed; the outcome for farmland is the same.

USDA’s Natural Resources Conservation Service operates the Agricultural Conservation Easement Program (ACEP), which pays landowners to permanently protect their land from non-agricultural development by separating the development rights from ownership. ACEP is one of the primary federal tools designed for exactly this problem. But its funding competes in the Farm Bill budget against other priorities, and it has never been funded at a scale that matches the pace of conversion pressure.

Agrivoltaics: A Partial Answer

One alternative model gaining attention in agricultural land-grant research is agrivoltaics — designing solar installations to allow simultaneous agricultural use beneath and between panels. Iowa State University Extension and Penn State Extension have both published work on dual-use configurations ranging from grazing sheep beneath fixed-tilt arrays to growing shade-tolerant crops in the alleys between panels.

Agrivoltaics doesn’t solve the problem when a developer is optimizing for maximum power density — which most utility-scale projects are. But it establishes an alternative model, and some states have begun incorporating agrivoltaic requirements into siting regulations for solar projects on high-quality agricultural land. The policy momentum is real, even if enforcement and uptake vary.

What agrivoltaics does is frame the choice differently: the land doesn’t have to stop being farmland to host solar. The challenge is getting developers to accept the added design and maintenance complexity, and getting state utility commissions and agricultural agencies to align on standards.

The Structural Squeeze

For family farms, what’s changed isn’t only that there are new buyers. It’s that the new buyers operate on capital timelines and with investment structures that most agricultural operations cannot match.

Private equity has been rolling up farmland on 7-to-10-year return cycles, extracting rent along the way. Data center and solar developers typically lock in land under 25-to-30-year agreements with options to renew. When a family weighing a farm estate — with heirs who may not want to continue farming and land valued well above any income it could generate in crops — receives an offer from an energy developer paying above-market for a long-term lease or purchase, the economics are hard to argue with.

The solution advocates and USDA researchers return to consistently is one that’s been accurate for 40 years: fund conservation easements and land trusts at a scale proportionate to the conversion threat. Corporate consolidation in the food chain shows what happens when financial capital captures the agriculture sector and extracts value without reinvesting in its long-term health. The energy sector’s emerging farmland appetite is the same dynamic with a different logo on the check.

The “data centers and solar farms” line in a farm business roundup looks like a footnote. Over the next decade, it may be the headline.