New Research Shows Many State Property Tax Laws Punish Fast-Evolving Technology

A National Taxpayers Union paper says state tangible personal property tax rules can overvalue data-center servers: Florida values a three-year-old server at 56% of original cost, while five other states value it at 40%, despite market value being close to nothing. The group argues these taxes can raise technology-service costs and calls for simpler, more uniform rules; only 7 of 38 OECD countries tax tangible personal property, and U.S. state definitions vary.

Analysis

Investment view: This is a slow-moving cost and siting issue, not a near-term earnings catalyst. If tax assessments materially exceed realizable server values, they can raise the after-tax cost of operating data centers and marginally weaken returns on refresh-heavy capacity. But the pass-through to customers is conditional: contract structure, utilization, local competition, and power availability may matter more than the tax line. For landlords, whether the burden reaches NOI depends on lease terms and who bears the tax.

Second-order effects: Over time, state tax treatment could influence where incremental capacity is built, but it is unlikely to displace power access, grid-connection timing, and permitting as primary location constraints. Reforming states could gain at the margin; states retaining aggressive schedules risk making marginal projects less attractive. The NTU paper is advocacy research, not evidence that operators’ reported tax expense is currently misvalued or that reform is imminent.

Timing and contrarian view: Any legislative repricing is more plausibly a 1–3 year catalyst than a days-long one; structural effects would accrue over 6–18 months as new capacity is sited. The market may overestimate the near-term benefit of reform: tax savings could be small relative to total project economics, while only a subset of assets and jurisdictions may be exposed. Reassess if filings quantify material exposure or states advance legislation; the thesis weakens if tax expense is immaterial or reform stalls.

Key Decisions for Investors

• No immediate sector trade: the article provides no company-level exposure, tax amounts, or evidence of enacted reforms. Avoid extrapolating the examples into an earnings estimate.
• Build a jurisdiction-by-jurisdiction watchlist for data-center operators and landlords, then verify assessed server values, tax expense, ownership/lease allocation, and facility concentration in filings before ranking exposure.
• Treat state reform as a conditional catalyst, not a base case. Consider a relative-value position favoring operators with verified exposure to states that enact lower or more accurate assessments only after legislation is passed and the savings are quantified.
• Falsifiers: company disclosures showing negligible TPP tax exposure, no meaningful tax allocation to the operator, or legislative proposals failing to advance over the next 1–3 years. Power and permitting constraints remaining dominant would also limit any location-driven earnings benefit.