- Tension: States sell data center tax exemptions as broad-based economic development, but the actual distribution of the benefit is concentrated in a way that undercuts that pitch.
- Noise: Headline investment figures and job-creation talking points dominate the coverage, crowding out the simpler question of who the money is actually flowing to.
- The Direct Message: A tax break built for an industry can quietly become a tax break for one company — and once that’s disclosed, the political cover for the policy disappears.
To learn more about our editorial approach, explore The Direct Message methodology.
Indiana’s data center sales tax exemption was pitched as a tool to attract a growing industry. New disclosures show it has functioned more narrowly: as a subsidy for one recipient, extracted from a pool of just six.
The exemption, and what it cost
Indiana offers data center operators a sales and use tax exemption on equipment and energy, available for up to 50 years for projects that invest more than $750 million. The program was built to compete with neighboring states for large-scale data center investment, and by that measure it has worked. Indiana has landed a wave of projects from major cloud and e-commerce companies, drawn in part by a subsidy structure among the most generous in the country.
The cost of that success is now on the record, though it took public pressure to get there. Between 2019 and 2025, the state disclosed $655 million in forgone sales and use tax revenue, according to a June 2026 report from Good Jobs First, a fiscal watchdog that tracks corporate subsidies. That disclosure did not come voluntarily. When Good Jobs First first flagged Indiana’s lack of transparency in an April 2026 study, the state comptroller’s office pushed back, arguing the losses were too small to be material — without defining what threshold it was using. Only after the watchdog group publicly challenged that claim did Indiana’s comptroller, together with the Indiana Economic Development Corporation, correct itself, concede the losses were material, and update its public transparency portal with the real numbers.
One recipient, most of the benefit
That $655 million went to just six companies. More than 93 percent of it — $611.5 million — went to one: Amazon. Amazon’s share grew from $50.5 million in 2024 to $561 million in 2025, an increase of 1,011 percent in a single year. No other company in the program comes close to that trajectory, and no other year in the program’s history shows that kind of jump for a single recipient.
That is not a program spreading benefits across a competitive industry. It is a program whose single largest line item grew tenfold in twelve months, concentrated in one balance sheet, out of a recipient pool that could fit around a conference table. When a state advertises a tax incentive as industrial policy, the implicit promise is that many companies will compete for the benefit, and that competition will do some of the work of keeping the program honest.
Six recipients isn’t a competitive market, and one recipient absorbing 93 percent of the money isn’t a market at all.
Indiana isn’t alone, but it is the sharpest case
The concentration problem is not unique to Indiana. The same Good Jobs First report found that Oregon’s Long-Term Rural Enterprise Zone program, a property tax abatement mechanism, has been dominated for years by exactly four companies: Amazon, Apple, Alphabet, and Meta. Between 2016 and 2025, those four companies collectively received more than $615 million in property tax abatements through that single program, with annual costs rising 762 percent over the period. In Oklahoma, a single Google data center complex in Mayes County received more than $363 million in property tax breaks between 2013 and 2025, with the annual abatement for that one complex alone reaching more than $40 million in a single year.
What sets Indiana apart is the speed and scale of the swing. Georgia’s data center sales tax exemption jumped 664 percent in a single revision, from an estimated $327 million to $2.5 billion for fiscal 2026 — a startling number, but one driven by an industry-wide undercount, not a single company’s balance sheet. Indiana’s 1,011 percent increase is different in kind: it is one company’s exemption growing more than tenfold in twelve months, inside a program that had already disclosed most of its numbers. The scale of the concentration, not just the size of the loss, is what makes Indiana’s case distinct.
The direct message
A subsidy is only as good as its distribution. Indiana built a program to attract an industry and got, in practice, one dominant beneficiary among six, whose share of the benefit grew by over 1,000 percent in a year. States weighing similar exemptions should ask not just what the incentive might attract, but who is actually cashing the check once the numbers come in — and whether they’d have found out without someone else doing the math first.
Why concentration changes the argument
The standard defense of data center tax exemptions rests on breadth: many companies investing, many communities benefiting, a rising tide that lifts a regional economy. Indiana’s numbers complicate that defense directly. When 93 percent of a subsidy flows to one recipient, the program stops looking like industrial policy and starts looking like a negotiated deal with a single, highly profitable counterparty — one that arguably didn’t need the incentive to build in Indiana in the first place, given the scale of its existing infrastructure investment nationwide.
That distinction matters for the legislature, which will eventually have to weigh $655 million in lost revenue against the jobs and infrastructure the exemption produced. Data center construction generates real, if temporary, construction employment, and the facilities themselves anchor some permanent operational jobs. But data centers are also, by design, capital-intensive and labor-light once built — the entire premise of the industry is that enormous computing capacity can be run with a comparatively small onsite workforce. A subsidy program built around job creation numbers looks very different once the primary beneficiary is a single company whose permanent employment footprint per dollar of exemption is modest.
It also matters for the public, which is being asked to accept the tradeoffs — higher electricity demand, land use changes, forgone tax revenue that would otherwise fund schools or infrastructure — without necessarily knowing that most of the benefit accrued to one of the world’s largest companies. And it matters that the disclosure only happened because an outside watchdog group forced the issue, rather than as a matter of routine government transparency. A state that has to be publicly pressured into admitting where $655 million went is not a state whose taxpayers can trust the next number it publishes without similar scrutiny.
What comes next
Indiana has not announced any structural changes to its data center exemption in response to the disclosure, unlike Ohio, which paused new applications to its own data center tax break within days of revealing that the program cost $1.6 billion in 2025 — twelve times the state’s original projection. Indiana’s program remains open, its terms unchanged, and its next major reporting cycle will show whether the 2025 concentration in Amazon’s favor was a one-year anomaly or the start of a pattern.
Given that the exemption runs for up to 50 years per qualifying project, a single year of data understates how long this concentration could compound.
