Written by HLRBO Staff|
Last updated
Data center sites made up just 3 percent of United States development land transactions in the first quarter of 2026 but 30 percent of the dollars, according to the commercial real estate firm Avison Young, and rural landowners are increasingly the people on the other side of those checks. For anyone holding ground that currently carries a hunting lease, the two options are not versions of the same choice. One is income. The other is an ending.
What the land market is actually doing
Avison Young, which brokers and tracks these transactions, reported that of the nearly 540 United States land sites that traded in the first quarter of 2026, data center sites "represent just 3% of all development transactions yet account for 30% of total dollar volume." Four of the ten largest land sales that quarter were planned as data center developments. In the firm's H1 2026 United States investment sales report, its analysts project total 2026 development site volume to reach $40 billion, which they call the highest mark since 2019, with data centers among the categories leading that growth across the 15 metros the firm tracks. That $40 billion is a top-15-metro figure, not a national one.
Treat those as what they are. Avison Young is a brokerage with a commercial interest in the category, and its numbers are its own market intelligence, not a federal dataset. They are worth reading as a signal about where capital is pointed, and not as a census.
The American Farm Bureau Federation's April 23, 2026 market intel piece, Balancing Data Center Growth with American Agriculture, puts the buildout at an estimated 4,925 active or under-construction data centers nationally, with Virginia at 706 and Texas at 546. Citing a Cushman & Wakefield cost guide, the authors note construction runs $9 million to $15 million per megawatt, which puts a typical 250-megawatt facility between $2.3 billion and $3.8 billion. That is the scale of money that decides a land price.
The number nobody puts side by side
Here is the arithmetic, with the inputs named so you can check it.
In March 2026, 96 landowners in Salem Township, Luzerne County, Pennsylvania closed on roughly 1,700 acres sold to Blackstone QTS for more than $500 million, as reported by Bill O'Boyle in the Times Leader on March 18. Divide $500 million by 1,700 acres and the floor price is about $294,000 an acre. The actual figure is higher, because the reported number was "more than."
Now the other side. University of Georgia Cooperative Extension Circular 971, Tips for Creating a Hunting Lease by retired wildlife specialist Michael T. Mengak, reports that "in 2022, deer hunting leases in Georgia were worth, on average, about $16.00 per acre (ranging from $10.00 to $30.00 or more per acre)." Mississippi State University Extension publication P2310, Hunting Leases: Considerations and Alternatives for Landowners, written by James E. Miller, Professor Emeritus of wildlife, fisheries and aquaculture, and reviewed in December 2023 by Extension Professor Daryl Jones, describes a range from "as low as 50 cents per acre" for small game up to "high-quality, big-game leases that may go for as much as $25 per acre or more in some areas." The same sentence puts something above that ceiling: Miller names "high-quality waterfowl blinds leases that bring the highest annual returns per acre of access," so $25 is the top of the big-game range rather than the top of the market.
Divide $294,000 by $16 and one acre of that Pennsylvania offer equals roughly 18,000 years of Georgia-average hunting lease income. Run it at $25, the top of Mississippi State's big-game range, and it is still about 11,800 years. For a third anchor, USDA's National Agricultural Statistics Service put United States farm real estate at $4,500 per acre in its Land Values 2026 Summary, released July 2026, with cropland at $6,020 and pasture at $2,000. The Salem Township floor price is about 65 times the national farm real estate average.
The conclusion is not that leasing is a bad deal. The conclusion is that these two things are not on the same axis, and anyone who tells a landowner to hold out for lease income instead of a serious development offer is selling something. A hunting lease is an annual return on ground you keep. A development sale is a one-time conversion of the ground itself.
The strongest case for taking the money
Two of the Salem Township sellers went on the record, and their reasoning deserves to be stated as they stated it. Doyle Watts, 61, sold 23 acres. "This deal gave me the opportunity to retire early," he told the Times Leader. Michael Hodgson, 37, sold 14 acres: "There were a lot of folks struggling out here. And now, their lives will be exponentially better."
That is not naivety. That is a rural economy answering a question it has been asked for decades without a good option.
The public finance case is just as real. Mike Turner, vice chair of the Loudoun County, Virginia Board of Supervisors, told an Agri-Pulse forum covered by managing editor Steve Davies on August 27, 2026: "We are going to get $1.3 billion in tax revenue from our data centers in fiscal '27." County revenue at that scale funds schools, roads and emergency services in places that have been cutting all three.
The Farm Bureau, which is hardly a booster, concedes the point in the same breath as its warning. Data centers, the market intel piece says, bring "increased local tax revenue, job creation and the potential revitalization of areas facing economic decline," and it opens by observing that "agriculture depends on data centers for innovation and efficiency."
What a lease is, precisely
A hunting lease is an annual contract for access. Title does not move. The timber keeps growing, the pasture keeps carrying cattle, the CRP contract keeps running, the mineral rights stay where they are. Next year you can raise the rate, change the lessee, add a second season, pull the ground back for family use, or sell to a developer anyway. The option survives.
That optionality is the actual product. A landowner who leases in 2026 has not foreclosed a 2029 offer. A landowner who sells in 2026 has foreclosed everything after it.
Lease income also behaves differently than sale proceeds on a tax return, in a divorce, in an estate and in a farm operating loan. That is a conversation for your own attorney and accountant, not for an article, and the rules differ by state.
What ends, and does not come back
Bernt Nelson, an economist at the American Farm Bureau Federation, is direct about both halves. On why farmers listen: "When a farmer is offered several times the value of agricultural land, it's understandable why they might consider having a conversation about that." On what follows: "It's also really important to think about what happens when productive farmland comes out of production." And on the scale of it: "Since 1982, we've lost about 56 million acres of productive farmland. That's about the size of Kansas, and this matters because productive farmland, once it's taken out of production, is pretty permanent."
The Farm Bureau piece Nelson co-authored with Autumn Lankford Higgins, the organization's director of government affairs, says the same thing in fewer words: "Farmland is the foundation of agricultural production and a generational asset for farm families. Once converted to industrial use, it is rarely returned to production."
For hunting specifically, the sale does something a lease never does. It ends the hunting on that ground permanently, and it does not stop at the property line, because a campus at the scale the Farm Bureau describes arrives with substations, transmission corridors, construction traffic and a permanent industrial footprint that the neighbors did not vote on. Those neighbors may be the people you have been splitting a deer lease with for fifteen years.
Neither Avison Young nor the Farm Bureau nor USDA publishes a figure for hunting acreage converted to data center use, and no one should pretend otherwise. What can be said is that the conversion is one-directional.
What is not known here
The Salem Township price is an outlier, and treating it as a benchmark would be dishonest. It reflects a specific corridor in eastern Pennsylvania with transmission capacity, fiber and a buyer with a balance sheet. Most rural acreage in this country will never see an offer remotely like it, and a landowner who declines a hunting lease while waiting for one is likely to wait a long time.
The lease figures are also softer than they look. Both extension publications describe rules of thumb from regional surveys, and Mississippi State says so explicitly. Nobody publishes a national hunting lease rate index the way NASS publishes land values, so the $16 and $25 anchors are illustrative of two Southern states in specific years, not a national average. Rates in Montana, Ohio and North Carolina are not the same number and do not move together.
Zoning, right-to-farm protections, conservation easement terms, agricultural assessment clawbacks and local siting authority all differ by state and often by county. Wherever your ground sits, the authority on what you can actually do with it is your county planning office, your easement documents and your own attorney, not a marketplace and not a broker's pitch deck.
A data center offer is a legitimate thing to take. Landowners have carried the cost of keeping rural ground rural for a long time, mostly alone, and nobody who has not farmed through a bad five years gets to lecture them about it. But the decision should be made with the right comparison in front of you. One option pays you every year and leaves the ground able to change its mind. The other pays once, and the gate never opens again.
If leasing is the direction, HLRBO's landowner tools are built for landowners who intend to keep title, and more reporting for landowners lives on the HLRBO blog.