Data centers lost the midterms

Saying data centers lost the midterms means voters and candidates pushed back against rapid, largely unrestrained expansion. The backlash appeared in election polling, local ballot measures, and several important victories by politicians who support moratoriums or stronger rules. It matters because data centers are a central part of the AI industry’s infrastructure.

The clearest example came from NBC News polling: 76 percent of Election Day respondents opposed data centers in their communities. Local voters addressed the issue in 45 ballot measures, and 33 results were expected to restrict development. Gina Hinojosa, Texas’s Democratic governor-elect, supports a statewide moratorium.

The results were not uniformly anti-development. Byron Donalds won Florida’s governor’s race while opposing an outright ban, and Rob Sand opposed a moratorium in Iowa. Still, even these candidates had to address the issue, suggesting developers will face continued pressure for limits and safeguards.

Data centers appeared directly on 45 local ballot measures nationwide. This shows the issue had moved beyond private planning debates and into local elections. It matters because communities used voting power to influence whether large facilities could be built, how large they could become, and how long development might pause.

According to a site compiled by Redbud Consulting, 33 of those measures had results that would restrict data center development as of Wednesday morning. Examples included a 180-day moratorium in Wells, Maine, and bans on facilities above specified power thresholds in several Ohio cities and El Dorado, Kansas.

The results point to broad local resistance, but they do not represent every election outcome or establish a national ban. Restrictions varied by place. Some were temporary pauses, while others barred certain projects for years or based on their peak electrical load.

Data centers are specialized buildings filled with computer servers, storage systems, networking equipment, and cooling systems. They process and store digital information. AI companies use them to train models, run applications, and handle user requests. The source identifies data centers as a central pillar of the AI industry, so limits on construction can affect how quickly that industry expands.

The article does not provide a technical definition, but its examples show what makes these facilities politically significant: projects can require very large electrical connections. Ohio communities considered bans for facilities above 25 megawatts, while El Dorado, Kansas, targeted sites with peak loads above 10 megawatts.

Their importance does not guarantee public support. NBC News found that 76 percent of surveyed voters opposed data centers in their communities. As a result, AI-related development is increasingly tied to local approval, energy planning, environmental concerns, and public demands for infrastructure costs to be covered.

Many voters and politicians fear that new data centers could place heavy demands on local resources and leave communities carrying the costs. The article specifically highlights electricity, water, grid connections, environmental impacts, tax exemptions, and the public share of facility profits. These concerns matter because data center expansion can affect utility planning and local development decisions.

Texas illustrates the conflict. Governor-elect Gina Hinojosa supports a statewide moratorium, while Greg Abbott ordered an audit that could slow new projects connecting to the state energy grid. In the Senate race, James Talarico emphasized environmental effects and public profit-sharing. Byron Donalds proposed making developers pay for their own energy and water needs.

The political response was not uniform. Some candidates supported bans or pauses, while others favored development with guardrails. Even candidates who opposed outright bans had to discuss the issue. That pressure suggests local concerns will continue shaping approvals, costs, and project design.

The restrictions ranged from temporary moratoriums to permanent or long-lasting bans on certain projects. Communities used facility size or electricity demand as triggers. Politicians proposed broader pauses, tighter approvals, and rules designed to make developers responsible for the infrastructure their projects require. These measures could slow expansion while officials assess local impacts.

Wells, Maine, favored a 180-day moratorium on large data centers, backdated to mid-June 2026. Several Ohio cities voted to ban facilities above 25 megawatts. El Dorado, Kansas, voted to prohibit data centers with peak electrical loads above 10 megawatts for 10 years. Pennsylvania’s Stacy Garrity supported a pause of up to three years.

Other proposals focused on payment and transparency. Byron Donalds introduced legislation requiring developers to pay for energy and water needs. Jon Husted introduced a bill requiring large facilities to cover infrastructure upgrades and energy costs. Josh Shapiro barred nondisclosure agreements for developments.

The election results could make data center construction harder to approve and finance. Moratoriums, bans, disclosure rules, and power thresholds create additional steps before projects can proceed. Candidates who won or gained support by opposing unrestrained development may face pressure to adopt guardrails, even where they rejected outright bans.

The main cost mechanism is infrastructure responsibility. Byron Donalds proposed making developers pay for their own energy and water needs. Jon Husted’s Ratepayer Protection Act would require large data centers to cover infrastructure upgrades and energy costs. Texas also announced an audit that could slow new developments from connecting to the state grid.

These policies could shift grid-expansion costs away from ordinary ratepayers and toward project developers. They could also delay connections or make some projects less attractive. The article does not quantify future construction costs or confirm which proposals will become law, so the exact national effect remains uncertain.

In general, data centers need large amounts of electricity because they run many servers continuously and must remove the heat those machines produce. Water may be used in cooling systems. The source does not explain these technical details directly, but it identifies energy and water needs as major political concerns. It also shows that some projects are measured by peak electrical load.

The article gives concrete thresholds. Several Ohio cities voted to ban data centers over 25 megawatts. El Dorado, Kansas, voted for a 10-year ban on facilities with peak electrical loads above 10 megawatts. Texas launched an audit that could slow new developments seeking connections to the state energy grid.

These demands pressure utilities to provide new connections and upgrade infrastructure. Proposals in Florida and Congress would make developers pay for energy, water, or grid upgrades. That could protect ratepayers, but it may also slow construction or raise project costs. The article provides no national estimate of the added demand.