The top line

38 of the 41 data center measures on Tuesday’s ballot had been called by Wednesday morning. Voters chose to limit data center development in 37 of them and to allow or support it in one. Three had not been reported.

Voters in 13 states faced at least 50 local ballot measures on data centers this year, according to Ballotpedia’s count. Forty-one of the 50 were on the November 3 ballot, in 10 states. Six were decided earlier in the year, and every one of the six went against data center development. Three more come in December.

No state had a statewide data center measure on the ballot. The one that was filed, a proposed Ohio constitutional amendment to ban data centers above 25 megawatts, did not collect its 413,487 signatures by the July 1 deadline. Its backers say they are aiming for November 2027. A Massachusetts initiative filed in August would need two-thirds voter approval and utility and water certifications before any local data center permit. If it clears review and signatures, it would appear in November 2028.

This post covers every measure we could find: what each one does, how it reached the ballot, and the result. Counts are as of Wednesday morning and will move as the last measures are called. Recall elections tied to data center decisions are at the end.

What was on the ballot

Of the 41 measures on November 3, a yes vote limited or conditioned data center development in 28 and allowed or supported it in 13.

Most of these measures came from residents, not councils. In Ballotpedia’s September count, 31 of 38 measures had qualified through citizen petitions and seven had been referred by a local governing body. The measures fall into four groups:

In the tables below, “Yes limits” means a yes vote restricts, pauses, or adds conditions to data center development. “Yes favors” means a yes vote allows a project or tells officials to support one. The outcome column records which way the result cut, regardless of how the question was worded.

Ohio

Ohio had 21 of the 50 measures, in 18 cities, villages, and townships. Voters went against data center development in all 19 that have been counted. Eleven bans passed, Perry’s financial guarantee rule and Grove City’s voter approval rule passed, Monroeville’s advisory question failed, Ashville’s EdgeConneX deal was repealed, and all four Wilmington rezonings were rejected by 80 to 20. The two Adams County township zoning votes had not been reported.

Ohio had more data center ballot measures than the next four states combined. Eleven of the 21 were bans, most of them citizen initiatives that amend a city or village charter to prohibit data centers above 25 megawatts. Wilmington alone had four referendums, each on a separate rezoning from Rural Residential to Light Industrial. Ashville voted on a term summary for negotiations with EdgeConneX. Two Adams County townships voted on whether to adopt zoning in their unincorporated areas, which would give them a say over land use, data centers included.

The Ohio measures share a background: the Ohio Capital Journal counts at least 138 municipalities and townships with data center moratoriums and 23 communities that have effectively banned them, against 257 data centers in the state. The statewide ban that missed this year’s ballot is aimed at 2027.

Every Ohio measure, by jurisdiction

Illinois

Illinois had 7 measures, all advisory. All seven went against data centers. The four Madison County townships and Joliet Township said no to permitting them, by 74 to 83 percent. Bloomington backed a moratorium with 78 percent and Morgan County backed restrictions with 88 percent.

Four of the seven were in Madison County, across the river from St. Louis, where Chouteau, Granite City, Nameoki, and Venice townships asked voters whether officials should permit new data centers. Joliet asked the same question about artificial intelligence data centers. Bloomington and Morgan County asked the opposite: whether to impose a moratorium or restrictions. None of the seven binds anyone. They matter because councils and county boards tend to cite them.

Every Illinois measure

Everywhere else on November 3

Thirteen more measures in 8 states: California, Florida, Kansas, Maine, Nevada, New Jersey, North Dakota, Wisconsin. Twelve of the thirteen have been counted, and eleven went against data center development. All three Kansas bans passed, all four Maine votes passed, and the Alhambra and Edgewater bans passed with more than three quarters of the vote. Janesville now requires a public vote on the GM site. Downe Township voters said no to an advisory question in favor. The one exception was Harwood, North Dakota, where voters rejected an added review requirement for industrial projects, 75 to 25. Boulder City, Nevada is still counting mail ballots.

Maine had four town votes, three of them moratoriums. Kansas had three citizen initiatives to ban high-impact data centers, including one in El Dorado that also covers battery storage. Alhambra, California and Edgewater, Florida voted on outright bans. Janesville, Wisconsin voted on whether any development of the former General Motors site costing more than $450 million must go to voters. Boulder City, Nevada had the one council-referred question that would open land to data centers, in the Eldorado Valley Transfer Area. Downe Township, New Jersey asked an advisory question in favor.

Every measure, by state

Decided earlier this year

Six measures were decided between March and September. All six went against data center development, by margins of 66 to 90 percent.

Two New Hampshire towns voted at their March and September town meetings. Monterey Park, California passed a ban in June with 88 percent. Augusta Township, Michigan voters rejected a rezoning for a proposed data center with 90 percent voting no. Wrightstown, Wisconsin voters said no, 87 to 13, to committing public resources to a large-scale data center. Port Washington, Wisconsin now requires a vote before the city creates a tax incremental district above $10 million.

Still to come

Clarksville, Arkansas votes December 1 on two bond authorizations for named data center projects, one for up to $50 billion in industrial development bonds for a Hatchbo LLC site and one for up to $5 billion for SF ARK1 LLC. Both come with payment-in-lieu-of-taxes agreements. Mandan, North Dakota votes December 15 on a 25-megawatt ban.

Measures that did not make the ballot

The 50 measures are the ones that survived. Ballotpedia tracked at least 17 local proposals in 13 states that did not reach a ballot, plus the Ohio statewide amendment. The reasons are a guide to where the next fights will be: signature counts, deadlines, court rulings that a zoning or incentive decision is administrative and not subject to referendum, and councils that repealed an ordinance rather than face a vote.

Recalls

Data center decisions have also produced recall elections. Ballotpedia counted 42 recall efforts against 80 local officials over data center projects through early October. On Tuesday, voters removed a council member in Mason, Michigan, the mayor of Yukon, Oklahoma, the mayor of Urbana, Ohio, and three Piqua, Ohio commissioners. San Angelo, Texas had not reported.

Recalls are the other half of this story. Where residents cannot put a project on the ballot, some have put the officials who approved it on the ballot instead. Independence, Missouri recalled a council member on September 1 over a 90 percent tax abatement for a Nebius data center. The Town of Turtle, Wisconsin removed its longtime chairman on October 6 after a dispute that included a $1 billion Panattoni data center campus. On Tuesday the pattern held. In Mason, Michigan, the recalled council member lost to a challenger backed by the anti-data center group, and that group’s slate won all four regular council seats. In Yukon, Oklahoma, the mayor lost his council seat with more than 80 percent voting for the challenger. Piqua and Urbana, Ohio removed officials the same night they banned data centers.