A stronger federal tax incentive is approaching an industry already moving into rural America. From January 1, 2027, the next Opportunity Zone cycle offers more favorable terms through qualifying rural funds. WIRED reported that Searchlight Institute identified more than 100 data-center projects potentially within eligible rural areas. That is an eligibility finding, not evidence that those projects have claimed a benefit.
The policy dispute concerns what taxpayers receive in return. Data centers require immense capital but do not necessarily employ large permanent workforces. Microsoft, Meta and Amazon told WIRED they do not use the program for their data-center investments; Google did not respond. Meanwhile, Senator Josh Hawley has proposed excluding the industry entirely, as some states reconsider their own incentives.
What the rural incentive actually changes
Created by the 2017 Tax Cuts and Jobs Act, Opportunity Zones are designated census tracts where qualifying investments can receive favorable federal capital-gains treatment. Investors place eligible gains into a Qualified Opportunity Fund, which must direct most assets into qualifying property or businesses. The One Big Beautiful Bill Act extended the program and improved rural-fund terms for the second designation cycle.
After a five-year holding period, the usual basis increase is 10%; a qualifying rural fund increases it to 30%. Raising the tax basis reduces the taxable portion of the originally deferred gain. It is not a 30% tax rate, a cash grant or a promise that 30% of construction costs will be reimbursed. A separate provision allows investors holding qualifying investments for at least ten years to elect exclusion of eligible appreciation from federal capital-gains tax.
For certain rural property, substantial improvements may need to equal 50% of adjusted basis, rather than the standard full 100%. This concerns spending needed to qualify an existing asset, not a blanket halving of taxes. Buildings, substations, cooling, networking, backup generation and servers make data centers capital-intensive projects, but those expenses do not automatically qualify every participant.
A qualifying rural fund must hold at least 90% of its assets in qualifying rural Opportunity Zone property. A project still needs the designated tract and the relevant business, property and fund conditions; rural location alone is insufficient. The rural definition excludes larger cities and adjoining urbanized areas. The benefit primarily belongs to eligible investors, whose tax positions may differ from those of developers, landlords, infrastructure funds and operators.
Potential eligibility is not evidence of uptake
According to WIRED, Searchlight compared rural tracts with a conservative database of fewer than 700 planned or developing projects and found more than 100 potentially eligible centers. Other trackers list more than 1,500 proposed or developing facilities nationally, suggesting a potentially larger universe, not proof that all would qualify. Taxpayer confidentiality limits public tracking of actual claims.
Microsoft said it does not use Opportunity Zones to buy or build data centers, while Meta also denied using the program. Amazon said it had not claimed the benefit and was not actively seeking land in the zones. Spokesperson Julia Lawless described land and access to talent as siting factors and said the company did not plan to add Opportunity Zone status to its criteria. Google's lack of a response establishes no participation.
An operator's statement cannot necessarily settle the position of every partner. Landowners, developers, utilities, lenders, tenants and infrastructure investors can participate through different contracts, and a technology company may lease capacity from another owner's building. That makes entity-level disclosure important; it is not evidence that every potentially eligible project uses such arrangements or that a denial conceals a claim.
The rural construction shift came first
Pew counted more than 3,000 operating data centers and over 1,500 projects in development. Only 13% of operating facilities were rural, compared with 67% of planned facilities. Its planned category includes proposed sites, construction and land held in reserve, so these figures do not describe a pipeline certain to enter service.
Of planned facilities, 39% were in counties without an existing data center. Three quarters were in the South or Midwest, with the South alone representing 48%. Many receiving governments therefore face their first negotiation with an industry whose electricity and land requirements differ markedly from an ordinary office or warehouse.
Training and serving larger AI models requires dense chips backed by power and cooling. Developers seek reliable electricity, fiber, expandable transmission, manageable land prices and room for multi-building campuses. Land assembly can be easier outside metropolitan areas. Resistance over bills, generator emissions, water and noise can also push projects outward without removing those impacts.
Tax preferences enter a migration already under way. They may improve returns rather than cause a project that would otherwise never happen. Whether the same development would occur at the same site without assistance—the counterfactual behind subsidy policy—cannot be established for every project from private siting negotiations. The incentive can amplify AI infrastructure demand without being its original cause.
Construction value and lasting local value diverge
Opportunity Zones aim to support growth and jobs in distressed communities, but the program imposes no general job-creation requirement on every qualifying investment. NCSL says a hyperscale facility can support more than 1,500 construction workers, while a national review it cited found an average of ten jobs per data-center project in 2022. These are different measures and periods, not a matched construction-to-permanent-job ratio.
Illinois illustrates the importance of thresholds. Its subsidized centers had to create 20 jobs; 22 of 27 qualifying facilities reported exactly 20, and only one reported more than 30 full-time positions. This does not prove no economic benefit. It shows that a large capital commitment can satisfy an incentive with relatively little continuing employment.
Construction contracts, supplier purchases, workforce programs and a stable property-tax base can matter greatly to counties weakened by population loss and limited investment. A carefully negotiated center can contribute revenue and infrastructure. But concessions on property, sales, equipment and electricity can reduce receipts below the impression created by a headline investment figure. Searchlight's Emily Kraschel told WIRED that capital investment alone does not guarantee jobs or a wider local economic boost.
The public cost also needs the correct denominator. WIRED cited the Joint Committee on Taxation's estimate of $40.9 billion over ten years for the broader Opportunity Zone expansion. That is not a data-center-only subsidy bill. Neither the estimate nor a project's construction value demonstrates what an individual community will retain.
Who pays for the infrastructure?
Around-the-clock campuses may require generation, transmission, substations or pipelines. Developer payment of the full incremental costs differs fundamentally from their inclusion in the general utility rate base, where households and small businesses may face higher bills. Lower direct water consumption can also involve higher electricity demand; no single resource metric resolves the trade-off. Noise, backup-generator emissions, farmland conversion and emergency planning further affect local value.
Federal and state policies are moving in different directions. NCSL reported dedicated data-center tax incentives in at least 38 states, usually sales-tax exemptions; 14 extend relevant benefits to electricity consumption and 11 offer statewide property-tax preferences. Several states have moved to narrow, suspend or reconsider benefits. Those decisions need not stop construction when power, land, fiber and customer demand remain decisive.
An investor's federal benefit can coexist with state exemptions and local property agreements, while no single government evaluates the full package. Residents may see only the concessions disclosed at local hearings. Ownership, operating entity, major tenant, expected load, water plans, tax terms and binding infrastructure commitments reveal more than an eligibility map. Possible benefits must remain distinct from confirmed claims.
Exclusion, conditions and the evidence still to come
On September 17, 2026, Hawley introduced the No Tax Breaks for Data Centers Act to remove data centers from qualifying Opportunity Zone businesses and property while preserving other eligible investments. It was a proposal, not an enacted exclusion. His announcement cited estimates that 14–17% of data centers sit in qualifying zones; that location estimate does not measure claims or use of the future rural cycle.
Hawley argued that large technology companies do not need federal help to build on farmland. His Republican sponsorship also shows why the debate is not a neat partisan divide: conservative communities can support AI investment while opposing cost shifting or confidential deals. A blanket exclusion is simpler than national rules for jobs, water, power and noise, but treats a project paying its infrastructure costs like one with weak community protections.
Conditional eligibility could instead require employment, disclosure, grid responsibility or enforceable community agreements. Enforcement would still have to address temporary compliance, splitting projects among entities and classifying infrastructure expenses. The existing program already relies on technical asset tests, holding periods and tax reporting. Industry supporters can also note that there is no public evidence of widespread use of the enhanced rural benefit before its new cycle starts; prospective rules are being debated before investment structures become entrenched.
NCRC advocates local planning goals and community-led conditions, recognizing that rural governments may lack specialized legal and technical staff. Responsibility is divided among federal agencies, states, utilities, counties and residents. Financial eligibility tests cannot by themselves establish whether a county has obtained a fair bargain.
The next evidence will be Treasury guidance and confirmed 2027 designations, followed by fund disclosures, financing documents and voluntary company reporting—not maps alone. Congressional hearings, additional sponsors or companion legislation could reveal whether Hawley's proposal advances; reporting conditions could also appear in later tax, energy or permitting laws.
At county level, agreements allocating substation, transmission, water, road, emergency-service and decommissioning costs will show what support buys. Permanent job commitments need to be reported separately from temporary construction work, with protection against electricity-cost transfers. The central question is not whether a rural county welcomes technology, but whether measurable, enforceable local value accompanies the public benefits offered to its investors.
Sources
IRS Opportunity Zone rules · Pew development data · NCSL state incentives · WIRED reporting · Hawley's proposed legislation · NCRC community-impact analysis · Google News aggregation entry, not an official policy source