US midterm candidates find something to agree on: Reining in datacenters

Among 74 selected gubernatorial candidates across 36 states, 43% supported banning or pausing new datacenter projects and 51% favored tighter regulation or conditional development; about 5% had no stated position. Support for bans was bipartisan but higher among the Democrats examined (60%) than Republicans (30%). In Texas, Democrat Gina Hinojosa defeated Greg Abbott and will take office in January; the article notes governors can influence regulation, energy policy, and permitting, but targeted restrictions do not necessarily oppose all datacenter development.

Analysis

The investable risk is not a sudden collapse in AI demand; it is a higher, less predictable cost and longer lead time to convert announced demand into energized capacity. Restrictions can shift value from greenfield pipelines toward permitted sites, existing campuses, and developers able to secure power and water before local opposition hardens. That supports a relative-quality bias within data centers, but does not immunize incumbents from tighter operating conditions or utility cost-allocation rules.

The second-order exposure is power infrastructure. If projects face siting or grid-access constraints, spending may migrate toward transmission, substations, cooling efficiency, storage, and on-site generation rather than disappear. Conversely, utilities that have committed rate-base investment to forecast data-center load could face prudence challenges if projects stall or costs are shifted to households. The political signal is notable because opposition appears locally salient across party lines; campaign spending by the AI industry may not offset concentrated community concerns.

Near term, candidate positions alone are not a reliable earnings catalyst: authority is fragmented across state, local, utility, and permitting bodies, and selected-candidate surveys are not enacted policy. Over 1–3 months, track post-election appointments, proposed moratoria, utility tariffs, and permit hearings in high-growth markets. Over 6–18 months, the key risk is a cumulative delay/cost premium that lowers project returns and pushes hyperscalers toward fewer, larger, pre-approved campuses. Contrarian point: broad opposition may produce targeted conditions—not blanket bans—while power demand still creates compelling incentives for states to compete for investment.

Key Decisions for Investors

• No broad short of AI infrastructure on this signal alone. Treat election rhetoric as a permitting-risk indicator, not evidence that announced capex will be cancelled; wait for enacted rules or project-level delays.
• Prefer a relative-value watchlist: established data-center operators such as Equinix (EQIX) and Digital Realty (DLR) versus less-proven greenfield development exposure. Initiate only if permit data show new-site delays while existing-campus leasing and power availability remain firm; avoid assuming either operator is insulated from local restrictions.
• Monitor power-chain beneficiaries, including electrical-equipment and grid-construction exposure such as Vertiv (VRT) and Quanta Services (PWR), as potential offsets if spending shifts from new sites to power delivery and efficiency. Confirm order trends and project mix before taking a directional position.
• Track regulated utilities and regional power markets for load-forecast revisions, interconnection queues, and data-center tariff changes. A rising burden on residential customers could trigger cost-allocation limits; that would falsify the assumption that forecast data-center load automatically supports utility capex and earnings.
• Escalate risk if states enact broad moratoria, permitting timelines materially lengthen, or hyperscalers disclose delayed/relocated projects; reduce the thesis if rules remain narrow and project approvals, power contracting, and leasing continue without deterioration.