GOP Candidates Break With Trump on Data Centers to Boost Midterm Odds

A Reuters/Ipsos June poll shows fewer than one-third of Americans approve of the rapid data-center buildout, with utility bills rising as much as 267% alongside the $725B construction rush. While GOP candidates are increasingly breaking with Trump’s pro-expansion stance ahead of midterms, the political pushback could slow permitting and increase regulatory risk for data-center expansion. Overall, near-term sentiment is cautious as energy-demand concerns and potential policy friction rise.

Analysis

The market is likely underpricing how quickly this turns from a cultural issue into a permitting-and-rate-base issue. The immediate loser is any asset whose valuation assumes frictionless load growth in contested jurisdictions: data-center REITs, local utilities that were leaning on AI demand, and hyperscalers if they have to absorb higher mitigation costs for land, power interconnects, and community offsets. The second-order effect is a geographic split: projects in politically sensitive suburbs get delayed, while Texas/Southeast markets with faster approvals and cheaper power capture displaced capex.

Over 1-3 months, the trade is not about demand disappearing; it is about timing and option value. Political backlash raises the probability that utilities, governors, and local commissions demand more behind-the-meter generation, longer studies, and richer community concessions, which pushes out revenue recognition for developers but can support vendors that sell self-generation, backup power, switchgear, and grid hardening. That favors names like GEV, ETN, and parts of the distributed generation stack more than pure landlords.

Contrarian view: consensus is likely treating this as a noisy election-season headline, but the real mechanism is a higher cost of capital for siting in constrained power markets. If hyperscalers keep capex guides intact and local approvals in Virginia/Georgia/Texas keep clearing, the selloff should fade; if state PUCs and midterm candidates keep using utility bills as a wedge, the re-rating can last 6-18 months. The key falsifier is evidence that political pressure changes actual interconnection timelines, not just rhetoric.

Key Decisions for Investors

• Short EQIX / DLR on rallies over the next 1-3 months; the setup is multiple compression from rising permitting and power-cost uncertainty. Cover if leasing spreads and pre-let activity stay intact through upcoming earnings and capex commentary.
• Long GEV or ETN vs. short DLR as a 6-12 month pair: if data-center projects face more local pushback, customers will spend more on on-site power, backup, and electrical bottleneck mitigation. Invalidate if order growth slows materially or if state-level approvals normalize.
• Reduce exposure to utilities with the most explicit AI-load optimism (AEP, PPL, NEE) until state-level rate cases and interconnection queues prove the political noise is not translating into slower load additions. Reassess if management raises data-center load forecasts at the next investor day.
• Watch item, not immediate trade: if midterm polling shows anti-data-center candidates gaining in VA/GA/AZ, expect a broader re-rating of AI-infrastructure beneficiaries; if pro-build candidates hold, the market may quickly reprice this as a transient headline risk.