How Tax Abatements Are Distorting Data Center Land Prices Across States

Two physically identical parcels in neighboring counties, with the same power capacity, can differ in price by 20 to 30% once a data center tax abatement is in play. The raw land is worth the same on both sides of the county line. What differs is the total cost of running a data center on each parcel over the 15-to-20-year horizon that most hyperscale and enterprise facilities are built around.

A property tax abatement of 50 to 100% for 10 to 20 years never appears as a line item in a land comparable. It appears in the bid price, and it explains why the county offering the abatement wins the deal.

For developers, site selectors, and investors who rely on raw land comps as their main market signal, this is a systematic blind spot. Land in a high-abatement jurisdiction does not trade at its raw economic value. It trades at a value that already includes the captured incentive. Any reading of land comp data has to start with understanding the incentive landscape of the target geography.

Why Do Tax Abatements Create a Land Price Gap Between Counties?

The mechanism is simple. A data center is a capital-intensive asset that operates for a long time. Over 15 to 20 years, the tax burden on a large facility can reach hundreds of millions of dollars. That includes property taxes, sales and use taxes on equipment, and, in some states, electricity taxes. An abatement that removes or reduces these obligations raises the project’s internal rate of return. Developers build that gain into their acquisition models and bid higher for the land.

The result is a price premium in high-abatement jurisdictions. On a raw comp it looks like stronger demand or higher land value. In a narrow sense it is, but the demand comes from incentive value, not from the land’s utility, its infrastructure access, or market fundamentals. A buyer in a no-abatement county who pays what looks like market price may be underpaying on a risk-adjusted basis. If the abatement next door is large enough to change the competitive outcome, that same buyer may be overpaying.

According to the National Conference of State Legislatures, at least 38 states have enacted tax incentives for data centers. Virginia’s program alone is estimated at $1.6 billion a year in foregone revenue, according to MultiState. That figure covers both lost sales and use tax receipts and the property tax abatements negotiated at the county level.

Which States Are Offering the Largest Incentive Distortions?

The size of abatements across the major data center states has grown sharply, according to Good Jobs First. Georgia raised its FY2026 cost projection by 664% to $2.5 billion in foregone revenue and expects nearly $3 billion in FY2027. Ohio lost $555 million in 2024 and an estimated $1.6 billion in 2025 to its data center tax exemption. Texas losses climbed from $1 billion in FY2025 to $1.3 billion in FY2026, and the five-year total for 2025 to 2030 is projected to reach $9 billion.

The property tax layer is where county-to-county price distortions show up most clearly. State sales tax exemptions apply evenly across a state, so they do not create price gaps inside it. Property tax abatements are different. They are negotiated locally by industrial development authorities, county commissioners, and school boards, and they vary widely within a single state. Two adjacent counties in Georgia, Indiana, or Texas can take completely different positions on abatements, and that difference is what produces the county-line price gap.

Many Georgia counties have negotiated property tax abatements of 50 to 100% for 10 to 20 years on data center projects. South Carolina offers fee-in-lieu-of-tax arrangements that cut property taxes by 40 to 70%. Indiana counties near the LEAP Research and Innovation District offer negotiated abatements, and these are a meaningful reason that area attracts gigawatt-scale investment. None of these are uniform state programs. They are county-level negotiations, and they create real pricing variation inside one state.

Here is how the main states compare in 2026:

Virginia offers a full sales tax exemption on equipment and electricity, with county-level PILOT agreements on property tax. Its estimated annual foregone revenue is $1.6 to $1.9 billion (FY2025 estimate). The incentive value is already embedded in land prices, so comps reflect it.

Georgia offers a full equipment exemption, covering the 4% state tax plus local taxes, along with county-level property tax abatements of 50 to 100% for 10 to 20 years. Its FY2026 projection is $2.5 billion in foregone revenue. County-to-county price variation is strong.

Texas offers a full exemption on equipment and electricity, and property tax abatements are negotiated by each taxing unit. Foregone revenue is $1.3 billion or more in FY2026. The impact on land prices is significant, and the Dallas-Fort Worth premium is partly incentive-driven.

Ohio offers a full exemption on equipment and electricity, with county-level property tax arrangements. New applications were paused in 2026. Foregone revenue was $1.6 billion in 2025. The land price impact is high, and the Columbus premium reflects earlier incentives.

Indiana offers a full equipment exemption, with county-level abatements near the LEAP District. Foregone revenue is not publicly disclosed. The effect is still emerging, with LEAP abatements pushing up land premiums.

North Carolina offers a full exemption for projects with a minimum investment of $75 million, plus negotiated local arrangements. Foregone revenue is not publicly disclosed. The impact is moderate, and part of the Research Triangle premium comes from incentives.

South Carolina offers a full equipment exemption, and its fee-in-lieu-of-tax arrangements reduce property tax by 40 to 70%. Foregone revenue is not publicly disclosed. Low baseline land cost combined with these arrangements gives the state a material advantage.

How Do Incentive Distortions Affect Land Comp Analysis?

The practical problem for site selection teams and investors is that land comp databases do not record the incentive context behind each sale. Take a comparable sale in Loudoun County, Virginia, and another in neighboring Fauquier County. They may look alike on a price-per-acre basis. But suppose the Loudoun sale came with a negotiated PILOT agreement that removed property taxes for 15 years, and the Fauquier sale had no such agreement. The two are not truly comparable. They carry different total-cost-of-ownership profiles that happen to look the same at the point of land purchase.

In 2026, the strongest data center sites are not the cheapest ones. They are the ones where the major constraints are clear early. A state with a bigger headline exemption can lose to a state with cleaner qualification rules if the uncertainty around the larger exemption makes it hard to model in the project’s financial plan.

Incentive risk is also becoming more relevant. Several states have moved to pause, cap, or reform their data center incentive programs in 2025 and 2026. Ohio paused new applications after its $1.6 billion single-year exposure became public. Virginia’s incentive is under active legislative review, and Arizona has seen cap proposals. A buyer who pays a price reflecting a full abatement, in a jurisdiction where that abatement faces political pressure, carries a residual risk. That risk appears in neither the raw comp nor the current zoning record.

What Happens When Incentives Are Withdrawn or Reformed?

Land in high-abatement jurisdictions is not priced on raw economics. It is priced on a total-cost-of-ownership model that includes the incentive. When the incentive is reduced, capped, or withdrawn, land prices adjust, but not always right away and not always in the open.

When a program is paused or capped, new deal flow moves to other jurisdictions faster than prices fall in the old market. The most dangerous window falls between the moment an incentive becomes politically uncertain and the moment the market has fully absorbed that uncertainty. During that period, existing comps still reflect the full-incentive valuation even though the incentive’s future is in doubt.

Industrial development authority PILOT agreements are often negotiated under non-disclosure agreements, which keep the terms out of public view while talks are underway. State economic development agencies and local authorities have wide discretion, so the specific incentive terms behind a recent comparable sale may be unavailable to outsiders trying to explain the land price, as America’s Plan notes. This opacity makes the comp problem worse. Comps fail to show the incentive context, and the incentive terms themselves may be legally shielded from disclosure.

How Acres Supports Data Center Land Analysis in Incentive-Distorted Markets

Land analysis in markets with large abatement programs means separating incentive-driven pricing from fundamental land value. That takes current, accurate transaction data at the parcel level. Acres gives data center development teams one system with ownership, transaction history, assessed values, zoning, environmental signals, and infrastructure proximity context for more than 150 million U.S. parcels. Teams get a complete, connected view of ownership, risk, and opportunity, instantly. They can also skip the manual multi-county property records pull that is otherwise needed to assemble a baseline picture.

Land Comps Without Incentive Context Are Incomplete

The 20-to-30% price gap between incentivized and non-incentivized parcels does not show up in a raw comp analysis. It becomes visible only when incentive terms are layered onto the transaction history, and in many cases those terms are not public. Site selection teams that treat raw land comps as the full picture are making acquisition and valuation decisions on incomplete information.

The better approach is to run the incentive analysis alongside the land comp analysis. Identify the incentive programs active in each target county. Model the 15-to-20-year total-cost-of-ownership impact of each one. Then read the land transaction data in that light. Parcels in high-abatement jurisdictions that trade at a premium to non-abatement neighbors are often correctly priced once total cost of ownership is the measure. Parcels in jurisdictions where the abatement is at political risk may be mispriced in the opposite direction.

Ready to build a land analysis that separates incentive-driven premiums from underlying market fundamentals? Pull transaction history, assessed values, and parcel context for any U.S. target geography on Acres.