How Communities Are Bargaining With Big Tech Over Data Centers
Communities are increasingly unsatisfied with blanket promises of jobs and economic growth when considering the siting of data centers nearby. Unlike the factories of past industrial eras, data centers’ effect on the permanent workforce of a region is modest compared with their land, capital, and power usage. The public case for these projects is rarely focused on permanent job creation, but rather on growing the tax base and boosting investments in construction, infrastructure improvement, and energy sector development.
Related Reading
Viewed this way, data centers can create real local value. But they force communities to think harder about power, water, roads, public services, and tax burden. These tradeoffs can look unfavorable; recent polling found that seven in ten Americans oppose data centers for artificial intelligence in their local area. Public opinion poses new challenges for elected representatives, who must balance the upsides and downsides of development. According to the National Conference of State Legislatures, even though 38 states now offer dedicated tax incentives for data centers, as of this year lawmakers in at least 28 of those states have introduced bills to amend or tighten those subsidies.
In debates over data center development happening across the country, communities are asking new, concrete questions: What, exactly, will they get in return? And who pays if the project creates costs that show up after the ribbon cutting? As resource economists in Wyoming, we’ve watched as data centers have become tests of whether and how local governments can negotiate the physical terms of the AI economy. In successful negotiations, both communities and companies have a clear bargaining template: Residents understand the terms being negotiated on their behalf, and developers get agreements clear enough to maintain public confidence over time.
In Wyoming, the arrival of data centers is no longer hypothetical. Cheyenne has become the state’s data center hub—one industry tracker counts 21 facilities operated by eight providers in the Cheyenne market, with more under construction. In April 2026, Microsoft announced its intention to purchase roughly 3,200 acres to develop a data center, adding to the company’s existing data infrastructure in Cheyenne. Meta is building an $800 million campus south of the city; Related Digital has broken ground on a campus anchored by CoreWeave; and the proposed Tallgrass AI campus would be designed to scale to as much as 10 gigawatts, using more electricity than all Wyoming homes combined. The boom is happening outside of Cheyenne too: Prometheus Hyperscale is advancing a flagship campus near Evanston and has identified a second Wyoming expansion site outside Casper.
This growth is not accidental. For years, state and local officials have treated data centers as targets for economic development, with policies to leverage the state’s land, power access, and fiber infrastructure. Wyoming offers sales tax exemptions for qualifying data center equipment; operates a Building Resilient Communities (formerly Business Ready Community) program that can reduce electric and broadband costs in exchange for investment and payroll commitments; and exempts certain large facilities in existing industrial parks from industrial-siting certification procedures. Communities are already seeing the benefits. Thanks to data center construction and electricity consumption, Cheyenne’s revenues from building permit fees were up 31% and electricity franchise fee revenues were up 48%, according to the city’s 2025 budget update.
As data centers have become controversial, debates over them often focus on whether communities will give a simple yes or no for development to proceed. However, the real deliberation happening in communities is more often around how to convert legal, political, and regulatory risks of data center development into valuable and reliable community assets. To build credibility on both sides of the deal, the local value of the development must be visible. The most durable agreements allow residents to point to specific wins that convey benefits for the community.
In debates over data center development happening across the country, communities are asking new, concrete questions: What, exactly, will they get in return?
One example of such tangible benefits can be seen in the deal that the city of Cedar Rapids, Iowa, approved with digital infrastructure provider QTS, which was anchored by a minimum $750 million investment in the community, in addition to a city-owned Community Betterment Fund with contributions of up to $18 million over 20 years. The city of Stillwater, Oklahoma, justified the benefits of its Google data center campus with payment in lieu of taxes (or PILOT) agreements to local jurisdictions. The state of Connecticut requires qualified data centers to negotiate host-municipality fee agreements, so local governments are not left empty-handed after long tax abatements.
Another place where benefits can be emphasized is on utility bills. In many local debates, an underlying question is whether households and small businesses are being asked to underwrite power infrastructure built for a data center. Utilities and regulators are converging around new kinds of rate design and contract language that make transparent who will be responsible for new costs incurred. In such arrangements, utilities can put conditions on very large consumers that prevent the costs of their use from being shifted onto other customers.
Wyoming has been ahead of much of the country on this. Black Hills Energy’s Large Power Contract Service tariff, developed with Microsoft, showed how a utility can serve an unusually large data center load without asking existing customers to absorb the risk. The original arrangement with Microsoft let Black Hills draw on Microsoft’s own backup generation during periods of high demand, which helped defer the need for new utility-scale generation. Today, the framework is broader, but the basic bargain is the same: Large customers must bring or pay for the power resources and grid upgrades their projects require.
In other cases, data centers are being leveraged as a capital trigger for long overdue grid modernization efforts. In Pima County, Arizona, the billion-dollar data center campus Project Blue is structuring a long-term power arrangement with Tucson Electric Power to protect existing ratepayers, fund grid upgrades, and add cleaner electricity sources. Commitments like these must be explicit and transparent otherwise “grid modernization” sounds like a euphemism for making residents shoulder the costs of the data center’s power consumption.
The real deliberation happening in communities is more often around how to convert legal, political, and regulatory risks of data center development into valuable and reliable community assets.
This leads to a second aspect of political durability: transparency. Public justifications for development need to hold up. It’s important for communities to be able to have a clear answer about whether a project is reducing a burden it created, reducing a burden that already existed, or both. Some communities have been better at transparency efforts than others. Even while operating under a nondisclosure agreement, Pima County built a public FAQ that addressed the location, neighborhood proximity, noise, traffic, tax revenues, wastewater, and ratepayer protection concerns related to Project Blue.
The experience of The Dalles, Oregon, shows why transparency matters even when the economics of a deal are defensible. The city required data center developers to pay for water infrastructure upgrades in exchange for holding water rates for data centers the same as other users. Officials said data center revenue helped hold a 2025 water-rate increase to 7% rather than 35%. But water quantities in the agreements were redacted as trade secrets, which fueled skepticism. The lesson is simple: It is not enough for data centers to pay their way if the community cannot verify the details.
Much local policy experimentation is now underway to build transparency into new data center regulatory mechanisms. Phoenix, Arizona, has shifted data centers into a special permit process that explicitly addresses grid reliability, fire risk, emergency response access, and noise. DeKalb County, Georgia, is considering rules for data center zoning that would require buffers for buildings near residentially zoned parcels, county parks, and transit stops, which would vary by project size and energy need.
A third aspect of building political durability is the art of bundling. The best local deals do not rely on a single payment or a single tax instrument, but instead bundle several kinds of credibility: fiscal commitments communities can see; power arrangements that protect other customers; cooling and water strategies suited to local conditions; operating constraints on noise and emissions; and mechanisms that enforce transparency and accountability. Bundling is successful when separate concerns are addressed on enforceable terms.
It’s important for communities to be able to have a clear answer about whether a project is reducing a burden it created, reducing a burden that already existed, or both.
In practice, many negotiations begin with some version of a community wish list. But that phrase undersells what is happening—communities are usually asking for very concrete things, and the lists can be important instruments for reducing polarization within a community. Residents may disagree about climate change or industrial policy, but they understand classrooms, emergency response, childcare, parks, taxes, and noise. As a result, these so-called wish lists also function as maps of benefits various constituencies must see in the bargain.
For example, in Barber County, Kansas, public questions around a proposed Google campus asked about school and park improvements, sound and light pollution, and the expansion of STEM programs, internships, emergency services, and daycare capacity. This is not an abstract list: Residents were asking what would change about the neighborhood environment once the project comes to fruition.
The community benefits agreement for Lancaster, Pennsylvania, is a good illustration of a full bundle because it addresses a soup-to-nuts version of concerns about the data center and the health of the community. The agreement combines clean energy and low-noise commitments, a cap on water usage, two $10 million community funds, and a system for processing complaints and public reporting.
One last aspect of building political durability into data center projects is recognizing that no two bargains will be identical. The water bargain in Wyoming is not the water bargain in Washington, and neither looks like the one in southern Arizona or central Pennsylvania. In fast-growing suburbs, the hardest questions may be noise, land use, emergency response, and tax revenue. In former energy communities, the bargain may center on whether new digital infrastructure can help replace a decreasing tax base or will just consume local resources.
Finally, an important consideration for every data center is what happens when the project ends. Communities are getting savvier about preparing for technological changes, a campus that underperforms, or the disappearance of large electricity consumers after utilities and local governments have already made commitments. Barber County asked publicly about decommissioning obsolete facilities, and Google pointed to a draft decommissioning agreement. Lancaster’s agreement includes a provision on decommissioning and e-waste management. Even when these tools remain new or untested, they point toward something important: Communities want more than a ribbon-cutting plan—they want an exit plan too.
In the end, the right question is not whether communities should be for or against data centers. It is what terms make these projects feasible. Wyoming’s example is useful because it is a pressure test of an energy-producing state with communities that are already accustomed to large industrial projects. We can observe an emerging policy architecture taking shape as the scale of projects strains ordinary planning categories. Looking ahead, we can see that the next version of the data center bargain may go beyond water, taxes, and electrical substations to include nuclear siting, workforce transitions, and the future of legacy energy communities.
If the bargain can be made legible in Wyoming, it will offer lessons for other communities across the country in negotiating the terms that work for them. The best host communities will not be the ones that ask for the least or the ones that reject every proposal on sight. They will be the ones that can write the deal down clearly enough that residents, utilities, regulators, and developers all know what counts as success.