On Election Day in early November 2026, American voters were asked 41 local ballot questions about data centers, spread across 10 states. According to preliminary reporting from Route Fifty and Futurism, opponents of new facilities won all but one of them. The exception was Harwood, North Dakota, where roughly 75 percent of about 430 voters rejected a charter amendment that would have required independent review and a public hearing for large industrial projects. Some of the measures were advisory rather than binding on local officials, and final tallies may shift slightly, but the direction is not in doubt.
For an industry whose growth plans rest on securing land, power and water, this is a notable data point. Until now the data center backlash has mostly appeared in zoning hearings and utility dockets. A ballot is harder to lobby around.
What voters actually decided
Ohio accounted for the largest share, with nearly 20 measures by Route Fifty’s count and 21 by Futurism’s, and opponents won every one. Route Fifty also reports that a statewide constitutional amendment to ban new data center construction in Ohio could reach the ballot next year. Elsewhere, defeats for developers came in California, Illinois, Kansas, Maine and New Jersey, with more votes expected in Arkansas and North Dakota.
The details are instructive. In Emporia, Kansas, voters approved a bar on “high impact data centers” with just over 60 percent, a category defined as facilities needing a peak load above 10 megawatts or a daily water draw above 100,000 gallons. Preliminary results in Alhambra, California suggested a citywide ban. Earlier in the year, Monterey Park, California voted overwhelmingly for a permanent ban. Secondary reporting based on Ballotpedia data put support for bans or pauses at 70 percent or more in 28 of the 38 measures that had reported by then; we could not verify that figure against Ballotpedia directly.
Exit polling points the same way. Futurism cites a CNBC exit poll in which 56 percent of voters called data centers a crucial or major factor in their choices, and Route Fifty reports a 56 percent figure from NBC exit data, adding that 64 percent of those voters backed Democrats. The two outlets name different networks, so treat the exact source with care. What they agree on is that the issue carried real weight.
Reading the result without over-reading it
It is easy to stretch this. A Fortune commentary by Mark Minevich predicts an AI slowdown and a wave of federal regulation when the new Congress is seated in January. Those are predictions and opinions, not reported outcomes, and the piece itself mixes the two. What can be said from the reporting is narrower. Local voters, in a self-selected set of places where a measure was put on the ballot, consistently chose to restrict or pause new facilities. That is not a national referendum on AI, and communities where the question never reached a ballot may feel differently.
The industry-side reading, quoted in the Fortune piece from AlixPartners’ Andrej Danis, is that “every fully permitted data center under construction just got more valuable,” and that “part of the US slowdown gets exported.” The first claim follows logically: scarcity raises the value of approved sites. The second is a forecast. If American siting gets harder, developers will look for places with available power, permissive rules and decent connectivity, and some of that demand could land outside the United States. The same piece cites analysts who say 10 to 15 gigawatts of planned capacity may never come online by 2027, an estimate it attributes to Dan Ives via Yahoo Finance, which we did not verify independently.
The practical point for the industry is that opposition has found a format that works. Water draw, electricity bills and the sense of having projects imposed through questionable deals are recurring themes in the coverage; Brookings senior fellow Mark Muro is quoted by Route Fifty describing the technology as being imposed “fast as possible and often through questionable data center deals.” Developers who negotiate community benefits, publish power and water figures, and avoid secret agreements are in a different position from those who do not.
What This Means for Philippine Founders
Treat this as a market signal rather than a local event. The Philippines has been courting data center investment, and if American sites become slower and costlier to permit, the country could look more attractive to operators seeking capacity. That cuts both ways. The same objections voters raised, about electricity prices, water and who pays for grid upgrades, apply anywhere power is tight, and Philippine communities and regulators can read the same headlines. For startups, the nearer-term implication is about cost and availability of AI compute: if capacity growth in the largest market slows, expect cloud and GPU pricing to stay firm, and plan product margins accordingly. Companies building energy-efficient inference, power management or site-planning tools may find a more receptive audience. Beyond that, nothing in the reporting speaks to Philippine projects specifically, so any claim of a direct windfall would be speculation.
Share this article