Essay Essay #9 June 16, 2026

The Pitch vs. the Bill

6 min read · 1,201 words · Data Center Resist Club

The Pitch vs. the Bill

Every data center proposal arrives with the same basic pitch: jobs, tax revenue, economic development. What rarely gets mentioned in that pitch is that the tax revenue figure is almost always calculated before subtracting the enormous tax breaks the project itself receives — breaks that, in several states, have grown so large and so fast that they now show up as line items threatening entire state budgets. This essay is about the other side of that ledger.

How Big Has This Gotten

The scale of foregone tax revenue tied to data center incentives has expanded faster than almost any state forecast anticipated. A 2024 legislative audit found total state, local, and regional revenue losses from Virginia’s data center tax exemption reached $928.6 million in a single fiscal year — a program that now equals 42 percent of all economic development incentive spending by the entire Commonwealth. By early 2026, that figure had grown further: Virginia’s own Annual Comprehensive Financial Report now puts the annual tax abatement cost at $1.6 billion.

💸
$1.6 billion
Virginia’s annual data center tax abatement cost as of 2026 — enough to fund the entire operation of the state’s Judicial Department

Virginia is the largest example because it has the largest concentration of facilities, but the pattern is national. Illinois lost $371 million to data center incentives in 2023 alone — a 628 percent increase over the prior year. In ten of the twenty states that disclose at least some annual cost data, data center subsidy programs now cost more than $100 million per year.

Texas offers perhaps the starkest illustration of how quickly these projections can go wrong. The state’s comptroller’s office projected its data center sales tax break would cost roughly $180 million for the entire 2027-2028 budget cycle — a forecast made just three years before the program ballooned into one of the costliest incentive programs in the nation, now losing the state roughly a billion dollars a year. Growth this rapid was, in the comptroller’s own framing, simply unforeseen.

What These Incentives Actually Look Like

Data center tax incentive packages typically combine several distinct mechanisms, and understanding each one helps explain why the total cost is so difficult for the public to track in real time.

Common Incentive Mechanisms

  • Sales and use tax exemptions: Eliminating tax on the servers, cooling equipment, and other hardware purchased for the facility — often the single largest component, since equipment is replaced and repurchased on an ongoing cycle
  • Property tax abatements: Reducing or eliminating local property tax on the facility itself, frequently negotiated at the county level for 10 to 20 years
  • Job tax credits: Per-employee credits, typically modest in total value given how few permanent jobs these facilities create, as discussed in Essay #1
  • Infrastructure subsidies: Public investment in roads, water lines, or grid upgrades that primarily benefit the new facility

A typical incentive package can reduce the effective cost of a $500 million data center project by $50 to $100 million or more — and because states compete directly with each other for these projects, a developer can simply take the same proposal to whichever state offers the better deal. Georgia, for example, offers a full sales and use tax exemption on data center equipment for facilities investing at least $100 million and creating at least 10 jobs, potentially saving a qualifying facility $20 to $60 million over the equipment’s lifecycle, on top of negotiated county-level property tax abatements running 50 to 100 percent for 10 to 20 years.

The Rural Bargain — and Its Limits

The original case for these incentives, particularly in Virginia, was framed around rural economic development. The incentive program began in 2008 with a modest bill intended to attract a then-unfamiliar industry to rural Mecklenburg County, Virginia — a county facing the kind of economic distress incentive programs are designed to address. That original rural framing is part of what makes the program politically durable even as its cost has exploded: proponents can point to genuine economic activity in places that needed it, even as the bulk of subsequent growth concentrated in suburban Northern Virginia, an area that did not share the same economic need.

This dual character — real benefit in some places, enormous and growing cost everywhere the incentive applies — is exactly why the debate over scaling these programs back doesn’t break cleanly along the usual political lines, and why even the legislators who created these incentives are increasingly the ones proposing to limit them.

The Reversal Is Already Underway

State legislatures are no longer treating these incentive packages as untouchable. Virginia’s own legislature has introduced bills to reduce tax incentives specifically for replacement equipment and repairs, and separate legislation conditioning continued incentive eligibility on facilities using emission-free backup generators — directly tying the tax benefit to the environmental performance covered in Essay #5. Georgia’s legislature has moved its own discussion about eliminating data center credits from a fringe position to a serious one. In Virginia specifically, the state Senate’s budget bill proposed ending the data center tax incentive entirely on January 1, 2027 — a direct contrast to the governor’s own competing proposal to extend the same incentive from 2035 out to 2050, illustrating just how contested this question has become even within a single state government.

Indiana has taken a different approach worth watching: rather than eliminating incentives, a proposed bill would require data centers to direct a portion of their sales tax revenue specifically back to local governments — an attempt to address the core community complaint that these projects generate revenue mostly captured at the state level while leaving the host locality to absorb the local costs.

“With all economic development projects, it’s generally a lot easier to measure the cost of the incentive directly versus the benefits.” — Policy analyst, on the difficulty of accounting for data center tax incentives, February 2026

What to Ask About Any Proposed Incentive Package

Most residents never see the specific terms of a tax abatement deal until well after it has already been negotiated and approved, often by a county board acting on a separate timeline from the zoning and permitting hearings discussed in Essay #7.

Questions Worth Demanding Answers To

  • What is the total projected value of the tax abatement over its full term, not just the first year?
  • Does the package include a clawback provision if the promised jobs or investment levels aren’t met?
  • Is there a sunset date, and has a vote already been scheduled to extend it before it expires?
  • Does the state require any public disclosure of the actual annual revenue loss, or is the figure only available in aggregate statewide reports years later?
  • Is any portion of the resulting tax revenue legally required to flow back to the host locality specifically, or does it go entirely into the state general fund?

The Bottom Line

The jobs-and-tax-revenue pitch behind nearly every data center proposal depends on the public never seeing the other half of the math: the tax revenue the state and locality give up to land the project in the first place. In state after state, that giveaway has grown far beyond what anyone projected, fast enough that legislators who created these programs are now actively working to scale them back. Communities evaluating a proposed facility have every right to demand the full incentive package in writing, with its total projected cost over its entire term, before accepting the economic development pitch at face value.

← Back to All Essays
Scroll to Top