Voters Turn Against Data Centers, and the Friction Lands at Site Selection

Voters rejected data center projects in Tuesday's midterm elections, and Democrats won control of Congress in the same contests. The New York Times reported the rejection and the Democratic sweep, and Time reported that opposition to AI data centers helped shape key races.

Data centers are warehouse-sized buildings full of servers that run AI models. The opposition was concentrated in local ballot measures. Reuters had reported on October 29 that voters in 10 states would have data center-related initiatives on their ballots on Election Day. Fast Company reported that voters chose to ban or pause local data center development in 38 of 41 ballot measures. Forbes reported that dozens of municipalities across at least seven states voted on bans or moratoriums, and the Wall Street Journal reported that several towns in Ohio, Kansas and Florida passed bans.

The same Journal coverage cited a poll of more than 79,000 registered voters, in which 56% said data centers were a crucial or major factor in their picks for Congress. Reuters reported that Democrats who promised data center moratoriums won state offices. A moratorium is a temporary legal halt, usually on approvals or construction. It differs from a permanent ban in how long a project can be stalled.

Equity markets moved in the same direction. Reuters reported that AI data-center stocks slid after the Democratic sweep, while renewable stocks rallied. The backdrop was already tense. Reuters had reported on October 22 that AI capital spending is a key theme driving the U.S. stock bull market, and that investors saw risk building in the AI trade as the data center backlash grew.

MarketWatch reported that U.S. data-center spending is not expected to drop significantly, but that delays, relocations and higher expenses for individual projects are expected.

The distinction matters more than it sounds. Total spending can stay flat while a flat aggregate line hides slower schedules at specific campuses, projects moved to other jurisdictions, and higher costs per project.

In my view, the friction lands at site selection. Local bans and pauses hit the step where developers have the least room to maneuver, because a rejected site cannot be fixed with a financing change. For investors, the question is less whether AI capital spending falls in aggregate and more whether the pipeline slips. That would show up in the timing of revenue for the companies building and supplying these facilities. For ordinary savers, the exposure runs through index funds and sector funds that hold those names, and through the utility bills of communities that host or turn away the facilities. Whether either effect becomes material depends on how many projects actually move.

Two cautions apply. A stock move on election night is a reaction to news, not a measure of lasting earnings damage. And this is one election, with one national poll and one set of local votes. The next round of permitting decisions and project announcements will say more about whether the results change the build-out than the vote tally does.