Decatur Township negotiated fiscal agreement — Indianapolis / Sabey
Indianapolis approved a Sabey abatement/community package in September: approximately $242 million in preferences and at least $29.8 million in promised community investment.

Established trend · Active
Latest assessment: · Evidence updated
Host governments and developers increasingly describe tax preferences together with retained taxes, community payments and public-service commitments. The mechanism is bargaining over the public return to scarce land, power, permission and services, rather than treating headline private investment as the public benefit.
Scope: Selected U.S. fiscal negotiations; approval, proposals and delivery distinguished.
Sabey and Brookwood illustrate negotiated packages; Lea and Eagle-type proposals expose unfinished terms, while Mount Pleasant changes incentive flows. Together they show a recurring bargaining practice. Whether it improves public leverage or net fiscal returns depends on the obligations governments secure and the costs they retain.
Indianapolis approved a Sabey abatement/community package in September: approximately $242 million in preferences and at least $29.8 million in promised community investment.
Brookwood/Applied Digital: reported September 24 approval combines about $314.5 million in abatements and $270 million in benefits over 20 years for a proposed $3.2 billion campus. They are different streams, not directly comparable net amounts.
Lea County opened consideration of a Google campus and up to $134 billion of industrial revenue bonds. The ceiling is not a county cash expenditure; the nearby proposed 2-GW solar development does not establish off-grid supply.
Microsoft chose to forgo up to $5 million annually under its 2023 Mount Pleasant agreement. The Village describes about $4.9 million for public uses and $100,000 for the economic-development corporation; final 2027-budget allocation and operative duration remain open.
Chesterfield proposed using prospective data-center revenue to reduce vehicle taxes.
Storey County/Nevada proposal would change local consent and accountability around data-center abatements.
Later approved evidence — October5: Proposed impact disclosure, public review and review-cost recovery; broad proprietary-information exception remains.
October 6 Indiana reporting describes the American Infrastructure Alliance’s campaign for uniform rules on energy costs, water and community benefits. Its coalition joins technology/infrastructure interests and construction labor. The Indiana campaign argues that developers should pay their own power costs and accept meaningful standards, while opposing moratoria. The reporting includes Kokomo project and future-revenue claims; those projections are not collected taxes or independently verified net public benefits. Coalition’s own proposals ↗.
This is political bargaining over which constraints development will face. Shared developer-union support can create a governing coalition while narrowing which public demands are considered legitimate. Construction-job support is not an enforceable worker-AI clause, and promised gross revenue is not a net public-return assessment.
Duke Energy and several large customers filed a negotiated settlement October 6; Duke announced it October 7. It would establish special electricity-service terms for new or expanded loads of at least 50 megawatts with consistently high utilization, or at least 150 megawatts regardless of utilization. Earlier service arrangements receive substantial protection from retroactive changes.
The proposal requires financial security for infrastructure and long contracts, with minimum durations generally reaching 10 or 15 years depending on project size and ramp-up. Early departure would trigger defined payments, including remaining infrastructure obligations and part of remaining minimum demand charges. The North Carolina Utilities Commission still must approve the settlement. Its proposed January 1, 2027 start is conditional. Environmental advocates dispute whether the protections adequately isolate other customers from costs.
The important mechanism is an enforceable allocation of risks between a very large customer and the wider utility system. A negotiated promise to protect ratepayers becomes meaningful only through the approved accounting, security and exit provisions.
Preserve the existing proposed public-return example: the replacement retains recovery on exit and the option to credit above-cost amounts to other customers.
On October 9, Senators Elizabeth Warren, Richard Blumenthal and Chris Van Hollen released an investigation of seven data-center companies: Amazon, Google, Meta, Microsoft, CoreWeave, Digital Realty and Equinix. Staff used responses to their December 2025 inquiries and interviews to examine electricity costs, tax preferences, disclosure and employment claims.
The report identifies an important distinction: paying for a company's own substation is different from paying for upgrades elsewhere on the shared electricity network that its demand makes necessary. Staff argue that company promises to pay their full costs can leave those wider costs with other customers. The report also examines confidentiality agreements, equipment-tax exemptions and incomplete accounting of permanent jobs. These are findings from a congressional investigation, not a comprehensive independent measurement of every project's net public costs. Primary report, including the evidence and qualifications ↗.
For The Race, this sharpens the question behind our existing utility and fiscal cases: who defines the costs that a developer must pay, and can the public inspect that definition? It provides evidence for testing bargains such as DTE's conditional protections; it does not establish that every arrangement fails.
A voluntary payment benchmark and locally administered fund make proposed public returns more explicit. The benchmark includes required grid-modernization spending; no received payment or net public gain is established.
Comparisons of cost allocation, contributions, transparency and remedies sharpen the public-return ledger. Contract provisions do not measure enforcement or net benefits; existing Lancaster/Frederick/Pima/Quincy cases are overlapping evidence.
PILOT and service investments combine foregone taxes with negotiated public returns; executed/net terms remain open.
Projected taxes test the gross-versus-net value of a specific public bargain.
Counterfactual state school funding tests who bears the cost of incentives; it is not measured district cuts.
Forecast payments and infrastructure contributions help explain the city’s fiscal stake, without establishing realized losses or guaranteed obligations.
Proposed incentive exclusion changes the public-return bargain by refusing county subsidy eligibility; it is not adopted fiscal policy.
A council appropriation identifies how existing tax receipts will fund host-neighborhood programs. This is evidence of a specified public return; delivery and net benefit are unverified.
Unclear for host governments over the contract term: final enforceable benefits, tax sacrifices and exit/security terms determine whether choices expand or harden.
Executed payment and cost-allocation terms, security against nonpayment, successor obligations and transparent net public-return accounts would show whether negotiation produces a durable public benefit.