The polling number that changes the site selection conversation
A UMass Amherst poll, led by political science professor Tatishe Nteta, found that only 11 percent of Americans would support an AI data center being built in their local community, while 65 percent oppose one. Nteta's framing is worth sitting with: he says this is an issue that could define a political career, and that opposition runs strongly across the political spectrum rather than splitting along the usual partisan lines. That last point is the one site selection teams have been slow to internalize. A project that clears every zoning and tax incentive hurdle can still face unified community opposition that has nothing to do with whether the local jurisdiction leans red or blue.
The scale of organized opposition backs up the polling. Ballotpedia counts 39 data center-related ballot measures across eight states this year, with Ohio alone responsible for 21 of them. Three Kansas communities, including Ottawa, will vote on outright bans. Four local ballot measures decided earlier in 2026 all produced overwhelming opposition to data centers. This is no longer a scattered set of NIMBY disputes. It is a recurring, organized pattern showing up on ballots across multiple states in the same election cycle.
Ohio as the epicenter, and why that matters
Ohio has become the clearest test case because it combines scale with political stakes. The state has 154 operational data centers and 84 more in the works, more than all but three other states. Nick Owens, an attorney with Conserve Ohio who describes himself as a lifelong Republican, told Stateline the state has shifted from welcoming tech investment to becoming what he called a "tech dump." Jen Belt of Protect Grove City said the issue has united residents across party lines and age groups in a way few local issues do, and Anthony Elmo of Good Jobs First argues that when the industry itself agrees to proposed regulations, that agreement is a signal the rules were too weak to begin with.
The politics have moved up the ballot. Ohio Republican Senate candidate Jon Husted is being attacked for championing data center tax breaks as lieutenant governor, with a National Republican Senatorial Committee memo reportedly calling data centers an "anchor" on his campaign. Husted has since drafted a bill requiring utility regulators to consider making data centers pay their own infrastructure costs, which Senate Democrats blocked, calling it political messaging rather than substantive policy. His Democratic opponent, Sherrod Brown, is running ads that make Husted the face of the data center buildout. Whatever the policy merits, a sitting candidate reversing position under this kind of pressure is the clearest signal available that the issue has real electoral weight.
The Texas moratorium breaks the red state assumption
The single most important data point in this story is Texas. Governor Greg Abbott was previously one of the data center industry's strongest state-level boosters, and he has now issued a moratorium on new data center permits. If the backlash were purely a progressive urban phenomenon, a Republican governor in a historically pro-development state would have no political reason to act on it. Abbott acting anyway tells CTOs running site selection models that red-state tax and regulatory friendliness is no longer a reliable predictor of how smoothly a project will clear local and state approval.
Pennsylvania adds a second data point in the same direction: both gubernatorial candidates, regardless of party, have distanced themselves from earlier support for data centers. Even national rhetoric dismissing opponents, Donald Trump has called them "backwards and poor," has not slowed the spread of local bans and moratoriums. Josh Vlasto, co-leader of the tech-backed super PAC Leading the Future, argues the backlash targets bad actors rather than the industry as a whole, and that well-run projects can still bring revenue. That argument may be correct, but it is now the industry playing defense rather than offense, and the polling suggests the public is not drawing Vlasto's distinction.
What this means for a multi-year capacity roadmap
If your company is planning a dedicated data center build, a large colocation commitment, or even a multi-year cloud region dependency tied to a specific geography, political risk now belongs in the same diligence category as power availability, water rights, and tax incentives. A site that looked uncontroversial eighteen months ago, when the local jurisdiction welcomed the investment, can become a ballot measure target before a project breaks ground, let alone before it reaches commercial operation. The planning horizon for a major facility, often three to five years from site selection to full commissioning, is long enough for the political environment to shift entirely during construction.
Practical diligence now means tracking Ballotpedia-style local ballot measure activity in target markets, not just current zoning status, and building contractual flexibility into site commitments so a permit freeze or late-stage local opposition does not strand capital in a single location. Companies and the Data Center Coalition notably did not respond to Stateline's requests for comment on this story, which itself signals an industry still deciding how to respond publicly rather than one with a ready playbook.
The decision this actually forces
Diversify site selection across multiple jurisdictions by default, rather than concentrating capacity in whichever market currently offers the best tax package, because the tax package is precisely the thing most likely to become a political liability for the local official who approved it. A multi-site strategy that looked like unnecessary complexity a year ago now looks like the more defensible position given how quickly a single market's political mood can turn, and the modest premium of running two smaller sites instead of one large one is cheap insurance against losing an entire buildout to a ballot measure you did not see coming.
The companies managing this well are treating community engagement and local economic benefit commitments, concrete things like local hiring, water recycling, and no special tax abatement, as a cost of doing business rather than a negotiating concession, because the polling data makes clear that the old playbook of quiet negotiation and incentive packages is no longer sufficient to secure a project's social license to operate. Build that cost into the project's financial model at the outset rather than treating it as a contingency, because the Ohio and Texas examples show the alternative is a far more expensive fight after construction has already started and capital is already committed to the site.
