Data centre opposition unseats 18 incumbents in US primaries and local elections
Officials who backed data-centre projects lost their seats, a reminder that siting depends on community consent as much as on power and fibre.
Opposition to data centre projects contributed to the defeat of 18 incumbent officials in recent US primaries and local elections, Data Center Knowledge reported on 27 August. The losses span state legislatures, county commissions, city councils and a utility regulator, and in several cases challengers campaigned explicitly against a named project.
In Utah, state Senate President J. Stuart Adams lost his Republican primary to Stephanie Hollist, and Box Elder County commissioners Lee Perry and Boyd Bingham lost theirs after backing the Stratos data centre. Perry told The Salt Lake Tribune he believed the data centre vote cost him his seat. In Missouri, four Festus city council members lost following a $6 billion project, and two Independence council members were also defeated. In North Carolina, four-term Edgecombe County commissioner Donald Boswell lost to David Batts, whose campaign signs urged a no vote on data centres, in the area around the Kingsboro project.
Virginia delegate Geary Higgins lost to John McAuliff. In Maryland, three Calvert County commissioners lost Republican primaries, and Frederick County council president Brad Young lost a Democratic primary after the county opened more than 2,600 acres to data centre development. Oregon state senator Janeen Sollman lost a Democratic primary in the Hillsboro area, and Georgia Public Service Commissioner Tim Echols lost re-election.
The report noted policy responses at state level: Texas Governor Greg Abbott has moved to slow projects and prevent costs shifting to other customers, Governor Shapiro of Pennsylvania added environmental and transparency requirements, and Governor DeWine of Ohio paused new data centre tax-exemption requests.
Why it matters
Power availability and fibre have dominated site selection, but these results show that local consent has become a third gating factor. Elected officials now face a measurable electoral cost for approving projects, which will lengthen approval timelines and raise the value of sites where zoning is already settled. Developers unable to explain bill, water and noise impacts plainly will find contested counties harder to build in.
Power availability and fibre have dominated site selection, but these results show that local consent has become a third gating factor.
Developer seeks approval
Rezoning, permits, land purchases or tax incentives go before a county, city or state body.
Officials vote on the record
Elected commissioners, council members and legislators approve the project in public session.
Residents organise
Concerns about electricity bills, water, noise and land use turn into campaigns and candidacies.
Primaries and elections
Opponents run on the issue; the report counts 18 incumbents defeated across several states.
State-level response
Governors and regulators add conditions, pause incentives or move to limit cost-shifting.
What you can learn from this
- Siting runs through local land-use law. A data centre needs land zoned for industrial or data-centre use, and if it is not, the developer must seek a rezoning or conditional-use permit from a county or city body. Those bodies are elected, hold public hearings and vote on the record. That makes every approval a political act, which is why the consequences arrive at the next election rather than in court. Appeals exist, but they test procedure rather than popularity, so an unpopular approval can be lawful and still end a career. Hearing minutes are public, so a vote is easy to attach to a name.
- Ratepayer cost allocation drives much of the opposition. Connecting a large load often requires new transmission lines and substations, and under many tariffs those costs are spread across all customers of the utility. Residents see higher bills without a direct benefit. Large-load tariffs that make the data centre pay for its own upgrades, and for capacity it commits to but does not use, are the mechanism regulators are turning to. Minimum-bill clauses matter because a data centre that reserves capacity and then delays construction can leave stranded costs behind.
- Cooling determines the water and noise footprint. Evaporative cooling consumes water to remove heat, while air-cooled and closed-loop liquid systems trade water for electricity and larger equipment. Chillers, fans and backup generators produce continuous noise that carries across flat terrain. The choice of cooling design is therefore a community relations decision as much as an engineering one. Designs that recirculate water or use dry coolers cost more to run in hot climates, which is why the cheaper option keeps being proposed.
- Tax incentives are the lever communities can withdraw. Many jurisdictions exempt data centre equipment from sales tax or offer property tax abatements to attract projects. Pausing or conditioning those incentives, as some states are now doing, changes the economics without banning the projects outright. This is why incentive policy is where backlash first shows up in law. Incentive changes can also reach projects already in the pipeline, which is why developers price policy risk into land options.
- Distributed capacity spreads impact. One argument in the report is that many smaller sites cause less concentrated strain on a single grid node, water source or neighbourhood than a gigawatt campus. The trade-off is higher cost per megawatt and more complex networking between sites. Whether that trade is worth it depends on how much approval risk is priced into a project's financing. It also changes who the neighbours are: many small sites mean many small hearings instead of one large one.
We teach this
How to use this in practice
- Read your local zoning code for data centres. Search your county or municipal code for the terms data centre, data processing or server farm, and note whether they are permitted by right, conditionally, or not at all in industrial zones. Done looks like: a three-line summary of what approval a project near you would need, which body would vote on it, and how often that body faces election. Note the setbacks and noise limits too, because those are the conditions an opposing group will cite first.
- Find a large-load tariff and summarise it. Utility regulators publish tariff filings; find one for a utility that has proposed rules for customers above a threshold such as 50 or 100 MW. Extract the minimum contract term, the minimum monthly bill and who pays for grid upgrades. Done looks like: a five-bullet summary you could explain to a neighbour. Compare the minimum term with the typical lifespan of a server generation, which shows why operators push back on long commitments, and write down what happens if the customer leaves early.
- Build a site scorecard. Create a spreadsheet with rows for power availability, water source, cooling type, noise setbacks, tax incentives, zoning status and community sentiment, and score two real or hypothetical sites from one to five on each. Done looks like: a comparison that makes the non-engineering risks visible alongside the engineering ones. Weight the rows by how often each one appears in reports of blocked projects, and see whether the ranking changes.
- Draw the stakeholder map for one project. Take a project mentioned in the report or one near you, and sketch the developer, utility, grid operator, county board, state regulator, governor and residents, with arrows for who approves, who pays and who votes. Done looks like: a single diagram where you can trace how a residents' concern reaches an elected decision. Mark which arrows carry money and which carry votes, since the report's central point is that both now flow through the same people.
Sources
- Data Center Backlash Reaches the Ballot Box — Data Center Knowledge
Our reporting is an original summary; full coverage is at the links above.
Don't just read about it — build it.
Square 1 teaches the skills behind the headlines, with every line of your work graded by AI. Find your starting point in 3 minutes.
Get your free skill report