Voters Opposed Data Centers in 40 of 41 Votes in 2026

By Grace Kim

Inc. reported that voters nationwide rejected data center development on 40 of 41 local ballot measures in the 2026 midterms. Harwood, North Dakota, broke the pattern by voting against added construction scrutiny, not by backing restrictions on the industry.

What did the 2026 midterms show?

Inc. reported that Democrats won control of both chambers of Congress and several governorships. Many candidates ran on pro-regulation platforms often centered on the rapid growth of AI and the related crisis of affordability.

Exit polls showed 76 percent of voters said they did not want data centers in their communities. The national mood was clearest at the local level: 40 of 41 measures went against development, while Harwood alone rejected a hearing requirement.

The ballot votes came last week, as debate over AI's growth overlapped with the affordability crisis. The measures remained local, but their timing linked data center regulation to a broader election fight over regulation and costs.

Takeaway: The election combined Democratic gains with broad voter resistance to new AI infrastructure.

Why did residents push back?

Opposition had been building in the run-up to the election. Town hall meetings across the country served as focal points for residents to voice concerns about planned facilities.

The grievances were concrete. According to Inc.'s account, residents raised three main issues:

• The potential environmental effects of data center development.
• The rising cost of electricity.
• Stories of deafening noise from data centers in their neighborhoods.

Beyond town halls, voters used local ballot measures to limit development. In Ohio, voters in 18 communities chose to slow down or outright ban data center development, the Ohio Capital Journal reported.

In California, various cities voted to ban development. In Maine, many communities passed a six-month moratorium on all new construction, extending beyond data centers alone.

The spread from targeted data center limits to a broad construction pause in Maine showed how local resistance could affect more than one industry.

Takeaway: Voters objected on environmental, electricity-cost and neighborhood-noise grounds.

Why did Harwood vote differently?

Harwood's ballot question would have required public hearings before large industrial projects, including data centers, could be built. About 80 percent of residents voted against imposing construction-review requirements, the Washington Post reported.

The small town's population was 845. Residents nevertheless hoped to attract the development on a 900-acre compound just outside town.

Mayor Blake Hankey told the Washington Post: "The proposed facility would provide $1.65 million in property taxes once complete."

That gave residents a measurable reason to prefer the proposed campus over a broader public-review process. Completion would unlock the $1.65 million tax estimate, while the location offered a way to draw the project away from nearby Fargo.

Takeaway: Harwood chose a specific tax opportunity over broader public hearings.

What has to happen next?

Harwood's vote rejected added review; it did not finish the project. Inc.'s account did not identify the developer, construction schedule or project completion date for the proposed 900-acre compound.

The next test is whether the compound moves from voter support to completion. The $1.65 million property-tax estimate depends on the facility being finished, not merely receiving local backing.

Harwood's exception was also narrow. National exit polls put 76 percent of voters against having data centers in their communities, while roughly 80 percent of Harwood residents rejected construction-review requirements.

For other towns weighing similar projects, the contrast sets three tests: attract a taxable project, address environmental, electricity and noise concerns, and avoid the backlash that carried anti-development measures in Ohio, California and Maine.

Harwood must now deliver the 900-acre project, turn its $1.65 million estimate into property revenue and retain voter support for that trade-off.