The Local Questions Behind The Votes Against Data Centers

🔍 Read the full analysis: The Local Questions Behind The Votes Against Data Centers on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on monitors, keyboards and dev gear

• Business-only prices and quantity discounts
• Tax-exempt purchasing
• Multiple users, one account, clear invoices

Create a free Business account →Free to register

As an affiliate, we earn on qualifying purchases.Hide

Voters in 46 of 47 local data-center ballot measures across 10 states supported or were leading toward restrictions, according to a public tracker’s unofficial results as of Nov. 5. The votes covered different proposals—from outright bans to temporary pauses and advisory questions—so the count does not mean voters made one uniform decision about data centers.

The tracker, Data Centers on the Ballot, is maintained by Redbud Consulting, a polling and data firm. It groups outcomes as “Restrict” or “Allow” even when ballot language differs. As of 12:30 p.m. Eastern on Nov. 5, it listed 47 measures with results; 41 counts were complete for election night and six were still being counted. The tracker reported 260,177 votes across the measures, with 72.8% on the Restrict side. Those figures are unofficial, not certified totals.

The tracker showed Harwood, North Dakota, as the one exception: voters rejected a proposal to review rules for large projects, with 105 voting yes and 325 no, according to the state’s unofficial results. Among the measures checked against official election sources were Hubbard, Ohio, where a ban on data centers of any size passed 2,622 to 802, and Boulder City, Nevada, where voters rejected a proposal to make city-owned desert land available for data centers, 4,947 to 2,064. The Tribune Chronicle also reported Hubbard’s ban passing 76% to 24%.

The widely reported figure of 45 measures reflected an earlier tracker snapshot on the morning of Nov. 4, when results were posted for 40 measures. The tracker later added a nonbinding question in Belvidere, New Jersey, and a charter amendment in Piqua, Ohio. Ballotpedia listed 41 measures before the election, using a different count that included fewer measures in Maine, Kansas and New Jersey.

Local Votes Shape Data Center Siting

The results point to local approval becoming a site risk for developers, especially where proposed rules require public votes or restrict zoning. In some communities, a future project could face a direct ballot test rather than being decided only through planning boards or city councils. The measures do not establish a single national policy, but they show that data-center proposals can become consequential local election issues.

The debate has practical stakes for residents and businesses. The source report says opponents commonly cite water use and energy demand, while supporters point to jobs and tax revenue. Ballot language in some places set limits in megawatts or gallons per day, connecting the broader debate to local infrastructure and land-use decisions. The election results do not, by themselves, establish how much water or power any specific project would use or what economic benefits it would deliver.

Many measures targeted proposed or future projects, rather than campuses already under construction. Their effects will depend on the text adopted, whether a measure is binding, and how local governments apply it. A strong vote against one specific land-use proposal cannot be read as a vote on every data center or every project configuration.

Products Worth Considering

Bose QuietComfort Wireless Noise Cancelling Headphones, Black

AmazonView Latest Price

Anker SOLIX S2000 Portable Power Station, 2,010Wh, 1,500W

AmazonView Latest Price

Bose QuietComfort Headphones (2nd Gen), Wireless Over-Ear Noise-Cancelling Headphones, Black

AmazonView Latest Price

As an affiliate, we earn on qualifying purchases.

Different Ballots, Different Restrictions

The measures collected by the tracker covered several kinds of decisions. Some proposed outright bans, including Hubbard’s ban on facilities of any size. Others set size thresholds, often around 25 megawatts, or created temporary moratoriums. Some asked voters to approve or reject rezoning for a particular site, while others would require public approval before a project could proceed.

There were also advisory questions, which do not have the same binding effect as an ordinance or charter change. The tracker lists 37 measures as binding and 10 as advisory. That distinction matters: a vote can register public sentiment without directly changing local law. In addition, different ballot wording means a “yes” vote can support a restriction in one place but have a different meaning elsewhere; the tracker’s Restrict/Allow labels are a method for comparing outcomes, not a substitute for reading each measure.

The totals also depend on what is counted as a data-center measure. Ballotpedia’s election-day list included 41 measures across 10 states, while the tracker reached 47 by counting additional local questions. A separate report described “40+ races across 13 states” by also including recalls and candidate races. These are differences in scope, not necessarily conflicting vote counts for the same ballot questions.

“Nothing on the page is certified.”

— Data Centers on the Ballot, the Redbud Consulting tracker

Counts and Legal Effects Still Pending

All results remain unofficial in the tracker’s Nov. 5 snapshot, and six measures were still being counted. Final totals could change as election officials complete canvasses and certification. The source report says it independently checked seven results against official election sites and found they matched the tracker; it could not verify all 47 directly. For the remaining measures, it relied on the tracker’s linked county sources and local reporting.

The full legal consequences are not uniform. The source material identifies 37 measures as binding and 10 as advisory, but the effect of each binding measure depends on its exact text and local implementation. The material provided does not lay out the complete language or legal status of every measure, nor does it establish whether any result will face a challenge. It also does not provide a final certified, comparable count for all places.

Certification and Local Implementation

Election officials must finish counting and certify results before the unofficial figures become final. Residents and project developers will then need to look to each locality’s adopted ballot language, ordinances and charter provisions to learn which restrictions take effect and when. Measures requiring a public vote could shape the approval path for future proposals; advisory questions may instead inform later action by elected officials.

The next useful milestones are certified vote totals, publication or implementation of the final legal text, and decisions on any projects affected by the measures. The tracker’s tally offers a broad snapshot of voter sentiment, but it cannot answer whether a particular data center will be built, how a local rule will be applied, or what infrastructure and economic effects a proposal would have.

Key Questions

Did voters reject data centers in 46 places?

Not in one uniform sense. The tracker counted restrictions as winning or leading in 46 of 47 measures, but the ballots included bans, size caps, moratoriums, zoning referendums, approval requirements and advisory questions. Some votes concerned particular sites rather than all data centers in a community.

Are the Nov. 5 results final?

No. The tracker described the results as unofficial. It reported that six counts were still underway, and certification by election officials had not been completed.

What was the exception to the restriction trend?

In Harwood, North Dakota, the proposal to review rules for large projects was rejected, with 105 yes votes and 325 no votes in the unofficial count cited by the source report.

Did every measure create a legally binding rule?

No. The tracker classified 37 measures as binding and 10 as advisory. An advisory vote can express voter sentiment without changing local law on its own; the effect of binding measures depends on their wording and local implementation.

Why do reports give different numbers of measures?

They used different counting rules and snapshots. The tracker initially showed 45 measures, then added two; Ballotpedia listed 41 using a narrower set of local questions. Another count included candidate and recall races alongside data-center questions.

Source: ThorstenMeyerAI.com