Don’t vilify data centers for the wrong reasons — just demand better ones

Pinal County, Arizona, supervisors rejected a large data center campus in August amid resident objections, with developers also losing similar votes this year in Florida, Virginia and North Carolina. Lawrence Berkeley National Laboratory estimates US data centers used about 17 billion gallons of water for cooling in 2023, potentially rising to 2–4 times that amount by 2028; water used to generate their electricity is roughly 12 times cooling use. The commentary urges site-specific public reporting and realistic power, water and community plans, while noting newer cooling systems can reuse water and that large campuses may draw as much electricity as a small city.

Analysis

The investable signal is a rising “time-to-power and permit” premium, not a near-term collapse in data-center demand. Local opposition can strand development capital or delay capacity even where national water-use figures appear modest; this shifts advantage toward sites with credible grid access, water plans and community consent. Existing powered campuses and operators with execution capacity may gain scarcity value, while speculative projects in water-stressed or politically contentious locations face higher delay and redesign risk. The bottleneck also redirects spending toward grid equipment, on-site power and lower-water cooling, though none is an automatic winner: backup generation can itself become a permitting liability, and closed-loop or air cooling must be assessed against site-specific power and operating requirements.

For NFLX, this is an indirect infrastructure risk, not evidence of a company-specific cost or capacity shock. The article offers no data on Netflix’s contract exposure, cloud costs or service impact; the potential channel is broader infrastructure scarcity eventually affecting availability or pricing, which is too weak here to underwrite a trade. The contrarian point is that public backlash may not reduce aggregate demand—it may consolidate growth among better-sited incumbents and increase the value of already-secured power. The near-term catalyst is project-level permitting and interconnection progress; over 6–18 months, measured water/power disclosures and actual generator runtime matter more than industry pledges.

Key Decisions for Investors

• No trade in NFLX on this article alone. Treat it as an indirect sector watch item; verify any material change in cloud/CDN costs, capacity commitments or service reliability before revisiting.
• Within data-center infrastructure, favor diligence on suppliers exposed to grid connections, power management and efficient cooling over speculative campus developers—but make positions conditional on order/backlog evidence and customer concentration, not the commentary itself.
• Track local permit outcomes, utility interconnection timelines, water-stress conditions and limits on backup-generator use. A pattern of delays would strengthen the scarcity thesis; approvals and on-time grid connections would weaken it.
• Falsify the thesis if developers continue commissioning capacity on schedule despite opposition, or if reported power and water use per unit of delivered capacity falls enough to ease local constraints. Require site-level measured results rather than company targets.