Voters Reject Nearly All of 40-Plus Local Data Center Ballot Measures

Voters in ten U.S. states rejected nearly all of more than 40 local data center ballot measures on 3 November 2026, preliminary results show. Route Fifty, reporting on 4 November, found opponents successful in "virtually all 40 ballot measures across 10 states." The lone reported exception was Harwood, North Dakota.

The count and the ten-state total come from the Data Centers on the Ballot tracker, run by Redbud Consulting, which Route Fifty cites and which can lag county election offices. Totals differ between outlets and will shift as counting finishes. Final numbers belong to the tracker, not to any news summary.

Where did opponents win?

Ohio recorded the most votes. Nearly 20 towns, cities and counties voted, and in every Ohio race data center opponents appeared to have prevailed. Two named localities stand out:

• Emporia, Kansas, where a prohibition passed with just over 60 percent of the vote.
• Alhambra, California, which was on track to ban data centers with many ballots still uncounted.

Illinois, Maine and New Jersey also hosted measures among the ten states. Route Fifty notes that some of the votes were advisory questions and not binding on local officials.

What happened in Harwood?

About 75 percent of 430 voters rejected a charter amendment that would have required independent review and a public hearing before large industrial projects, including data centers, were adopted. The North Dakota secretary of state's office supplied the figures, as cited by Route Fifty.

The amendment would have applied only to future projects. The campus already approved in Harwood was exempt. The scale of the vote is small. The signal is not.

How many measures were expected beforehand?

Ballotpedia News, in a 21 September 2026 pre-election count, tallied 38 local data center measures across 35 jurisdictions in 30 counties in eight states, with 34 scheduled for the 3 November general election. Ohio led with 20 measures across 17 jurisdictions.

Of the 38, 31 qualified through citizen petitions and seven were referred by local governing bodies. Nine were advisory questions. Post-election counts exceed that pre-election figure, and the two tallies have not been reconciled.

Why does a land-use vote matter for infrastructure planning?

The sources report ballot outcomes only. They do not address specific industries beyond data centers. But the pattern carries weight for any capital project that needs rezoning, charter changes or development agreements, and for organizations weighing where such facilities can be built.

Three implications follow from the results:

• Siting risk is local and political. A facility that requires local approval can be stopped by a local vote. Planners of new capacity should check whether their jurisdiction can put such questions to voters.
• Contract terms matter. Organizations negotiating long-term capacity commitments may want provisions that account for delayed or relocated facilities.
• Community engagement is part of the plan. Route Fifty's coverage of exit polling notes that NBC polling found 56 percent of voters called AI and data centers a crucial or major factor in their vote.

Nothing in these results changes any funding or regulatory rule at state or federal level. This is a local land-use story.

What decides what happens next?

Certified final counts. Provisional tallies will move as county election offices finish, and the Redbud Consulting tracker is the reference for outcomes. The measures that were binding will take effect as local law; the advisory ones will test whether officials act on them. The next round of petition drives and referrals will show whether 3 November was a peak of local resistance or a baseline.