OH ballot / IN dead / ME veto

Why it matters

Three updates and a structural read. (1) The Ohio ballot initiative is the highest-blast-radius single regulatory event tracked in this newsletter to date. If 413,488 signatures are submitted by July 1 *and* validated, *and* the measure passes by simple majority on November 3, Ohio's constitution will prohibit any data center >25 MW — which by definition removes Ohio (currently the 6th-largest US data-center market by IT load, anchored by AEP territory's New Albany / Columbus cluster, plus emerging Lordstown / Mahoning Valley sites) from the hyperscale-eligible state list. The ~$30B+ of announced Ohio data-center capital expenditure (AEP's queued Lordstown loads, Microsoft New Albany, AWS Hilliard, Cologix Columbus) could be partially or fully stranded. The base-rate probability is low — citizen-initiative success requires (a) signature validity rates ~70-80%, (b) county distribution discipline, (c) survival of legal challenges, (d) majority approval on Nov 3 — but each step has happened at the rates required, and the Ohio Capital Journal's reporting suggests volunteer signature operation is meeting the per-county distribution targets. Cliff's regulatory-risk module needs a 'state-level constitutional ban in active signature collection' tier with Ohio at the top of the watch list, and any Ohio site under active diligence (especially Lordstown / Mahoning Valley which is also under a separate municipal moratorium covered Wed/Thu) needs an Ohio-specific risk overlay through November 3. (2) Indiana SB 257's no-hearing death is the *opposite* signal from Maine's veto-and-advisory-council death. Maine's death came with an executive order setting up a study commission — Maine is *actively* moving even after the moratorium failed. Indiana's death is silent — no hearing, no committee debate, no follow-on study commission, no governor-level action. Indiana's posture remains 'rate-increases-from-DCs-pass-through-to-residential-ratepayers' which is the structural opposite of Wisconsin's We Energies VLC tariff, Ohio's pending AEP rate case, and Florida's SB 484 just-signed cost-of-service mandate. The strategic implication: Indiana is the last large data-center market in the eastern US where the legislative status quo still favors developer-cost-shifting onto residential ratepayers, which makes Indiana the highest-cost-arbitrage state for hyperscaler siting — but also the highest-political-instability state if Indiana's legislature makes a sharper reversal in 2027. (3) Maine's advisory council pattern (15 members, report due early 2027) is the new template — a way for governors to preserve no-immediate-moratorium policy without giving up pressure-relief for environmental coalitions. Watch for the same template in Pennsylvania (active hyperscale buildout, active environmental opposition), Massachusetts (already has a small-scale framework), and possibly Michigan (active Sault Ste. Marie / Lansing siting). Cliff's regulatory-risk module should add a 'governor-level study-commission active' tier separate from 'moratorium active' — same eventual policy effect, very different short-term submission-readiness implications. The structural read across all three: state-level moratorium and tax-incentive activity is *highly* path-dependent, with the same 'ban data centers' outcome reachable via at least four different procedural routes (direct moratorium, ballot initiative, no-incentive-bill-passes, study-commission-recommends-moratorium), each with different timelines and reversibility. Cliff's regulatory-knowledge graph needs to model the *route* not just the *destination*.