Date: 2026-08-03. Synthesized directly from §2–§10 by the primary session. Depends on §10's scorecard; tests the protocol's seed hypothesis H10.1/§11's own premise (rulemaking-executable items first, cheap and slow-burning, before statutory/political items).
Tier 1 — rulemaking-executable now, already in motion (0–1 year)
TIER 1, ITEM 1 RE-SPECIFIED 2026-08-10. Two passes moved it. Phase 1 (verification-log.md, findings 1–2) established that the November 2024 document is OIG's statutorily mandated annual solicitation, does not mention CM, and that the actual CM proposal (RIN 0936-AA13) sits on Long-Term Actions with a July 2027 NPRM target — so "already underway" and "0–1 year" were both wrong. Phase 2 (steelman-log.md) established what the right item is: CM coverage through CMS §1115 demonstration authority, which five states have already obtained (California, Washington, Montana, Hawaii, Delaware) at approved incentive maxima of 596–1,092, with California delivering since March 2023. Application to approval ran about a year; approval to service, nine months. That is a real Tier 1 item with a demonstrated clock. The federal safe harbor is a Tier 3 item wearing a Tier 1 label.
Contingency-management safe-harbor expansion (A3). §6 confirms OIG has clear, pre-existing statutory rulemaking authority (Pub. L. 100-93 §14) to create a dedicated CM safe harbor or raise the nominal-value threshold broadly — and the process is already underway (OIG's November 2024 solicitation for new/modified safe harbors, comments closed January 2025). This needs no new appropriation, no new statute, and scores top-or-near-top under three of four §10 objective weightings. This is the filing's single highest-priority, lowest-cost recommendation.Replaced by: state Medicaid coverage of contingency management via §1115 (A3). The executable, evidenced, already-running path. It is not free — what states stand up is a staffed service (a per-site CM Coordinator, 36 point-of-care tests per patient, an incentive-management platform, a readiness review and ongoing fidelity monitoring), which is why §10's "near-zero net new cost" cell is flagged in the Phase 2 sensitivity. A dedicated federal safe harbor remains desirable — it is what would let voucher-magnitude protocols (~$1,536 per 12 weeks, per Rash et al. 2025) reach payers nationally — but on the published agenda it is a multi-year item and it does not gate the state route.- Medicaid IMD 1115-waiver expansion (part of A2/A4). §6: 38 states + D.C. already have an approved SUD-related waiver; CMS can approve more via existing authority. Cheap relative to new appropriations, though the ceiling on how much capacity this can unlock is set by the underlying exclusion Congress would need to remove for a durable fix.
- Settlement-fund spending-condition reform (from §7's H7.1 finding). Not a new architecture per se, but a governance fix on money already appropriated and flowing — improving public reporting requirements and evidence-basis conditions on the ~$50-58B in opioid settlement funds is arguably cheaper and faster than any architecture requiring new money, since it redirects existing disbursement rather than seeking new appropriation. §7 found only 12 states currently commit to detailed public reporting; closing that gap is itself a rulemaking/administrative-conditions problem, not a legislative one in most states.
Tier 2 — legally mixed, meaningful progress possible without full statutory fix (1–3 years)
- MOUD-everywhere via existing channels (A2). §4's X-waiver natural experiment is the key caution here: legal-access expansion alone (ED-initiation protocols, telehealth permanence) will not by itself move patient volume the way advocates hope — provider willingness, training, and payment are the binding constraint per two independent studies. Sequence this after a companion payment/training investment, not as a standalone legal fix.
- Cannabis's narrow-lane progress (A7, partial). §6: the April 2026 Schedule III narrowing (state-licensed medical products only) is done via rulemaking, but is under active D.C. Circuit challenge and does not touch banking, recreational status, or interstate commerce. Treat as partially landed, with the larger fix (SAFE Banking, full descheduling) queued in Tier 3.
Tier 3 — requires new statute, longer horizon (3+ years, or contingent on political conditions)
- Methadone deregulation (A1). §6: the OTP-only requirement is written directly into 21 U.S.C. §823(h); MOTAA exists precisely because it would need to amend that statute. This is the architecture with the strongest peer-country comparative case (§4: Australia, Canada, UK all dispense at pharmacy scale) but the least executable path in the US absent congressional action.
- Cannabis's full fix (SAFE Banking, descheduling). §6: SAFE Banking has passed the House seven times without a Senate floor vote across three Congresses — a specific, well-documented legislative blockage, not merely "needs more time."
- A durable/universal IMD-exclusion fix. §6: possible (SUPPORT Act 2018, CAA 2024 precedent exists), but each prior expansion required a dedicated legislative vehicle.
The public-order/political-durability constraint runs on a different clock than the legal-mechanics clock
§8's finding is the sequencing-relevant complication the legal-mechanics ranking above doesn't capture: both Oregon's and San Francisco's 2020-era harm-reduction-forward policies were reversed within 2–4 years by locally-organized, donor-backed coalitions, independent of how the underlying mortality-data disputes eventually resolved (§9's Oregon findings; §8's SF findings). This means any Tier 1/2 item that is politically legible as "harm reduction" (A5, A6) carries a durability risk that a rulemaking-executable, evidence-first item like CM (A3) does not — CM is not coded in public discourse as either "soft on drugs" or "harm reduction," which is likely part of why it has moved through the federal rulemaking process with comparatively little political friction while decriminalization fights consume years of coalition energy for outcomes that got reversed.
THE INVISIBILITY RATIONALE DID NOT SURVIVE PHASE 2 (steelman-log.md, Steelman C — the one steelman that survives unqualified). Low political visibility is scored here, and in §10's A3 D6 cell, as a durability asset. The record says the opposite: a policy nobody can see has nobody demanding it, and on a crowded agenda salience is what buys calendar time. Between 2022 and 2026 the federal government made four legally visible, contested changes to drug treatment and drug law — the X-waiver repeal, OTC naloxone, the 42 CFR Part 8 methadone reform, and the April 2026 cannabis Schedule III order — and all four happened. The one politically invisible item in this filing's own catalogue, a dedicated CM safe harbor, is the one that did not, and its target moved away from the present (May 2026 → Long-Term Actions → July 2027) across four consecutive annual OIG solicitations. The re-score parked the related objection — "OIG-reversibility is a different vulnerability than backlash" — in basis text. It is a different vulnerability and a worse one: a safe harbor made by notice-and-comment can be narrowed by notice-and-comment, and an invisible rule has no constituency to defend it. Low backlash risk and low durability are the same fact viewed twice. The comparison this section draws is also mis-specified: Oregon and San Francisco were ballot-and-ordinance policies reversed by electorates; the comparison class for a federal safe harbor is other federal safe harbors.
Sequencing implication: front-load Tier 1 items precisely because they are both cheap/fast and politically low-visibility — by the time any backlash coalition organizes (§8's pattern took ~2-4 years to reverse Oregon/SF), a CM safe-harbor expansion or IMD-waiver approval will likely already be delivering results, banked before opposition mobilizes. Decriminalization-family architectures (A6), by contrast, should not be sequenced as a first move even where the mortality/liberty case is strong (§10's scorecard shows A6 tied for best under liberty-first weighting) — both real-world attempts at this exact architecture were reversed within the political-durability window this project has now documented twice.
Kill-condition-adjacent finding for the whitepaper
Per §3's findings, the current overdose-death decline (~37-38% peak-to-2025) is most credibly attributed to supply-side/exposure-side mechanisms outside any single policy lever's control — not to any architecture this filing could recommend adopting faster. This means the status-quo-drift comparator (A11) is a genuinely strong baseline to beat, not a null strawman — every architecture above must be justified against a counterfactual where deaths are already falling substantially for reasons unrelated to new policy, a caution the whitepaper's honesty box should carry prominently, mirroring how GBMT-2's housing filing had to reckon with its own not-fully-attributable trend lines.
What remains before this can be presented as final sequencing
Per M6, this sequencing synthesis inherits every caveat §10's scorecard already flagged — no independent second scorer, no written red team, and dimension/tier weightings set by synthesizing judgment rather than pre-registered before evidence collection. Treat Tiers 1-3 above as a defensible first-pass structure, not a locked recommendation.