# Responsible America — Public Scorecard Methodology Status: initial data-model and methodology document, adopted this milestone. Companion to `docs/fact-checking-policy.md` and `docs/legal-review-policy.md`. Last reviewed: 2026-08-10. ## 1. What this is Division 17 of the platform's own charter ("Responsible America Public Scorecard") is described as existing "to close the loop back to the ordinary citizen" — tracking what actually happens to a proposal after it is published, not just what the proposal says. This document defines that tracking's data model and computation rules before any page renders a single disposition, so the rules are fixed and auditable rather than invented ad hoc per reform. ## 2. The eight dispositions Every reform on this platform is classified into exactly one of eight states, in the order a real proposal would move through them: | Disposition | Meaning | |---|---| | `proposal` | Published on this platform; no matching real-world enacted legislation has been identified. | | `enacted` | A real bill implementing this reform has become law (per `current_legislation[].status`). | | `implemented` | Enacted, and the responsible agency has taken the specific operational steps the law requires (e.g. a rule is finalized, a program is running). | | `measured_result` | Implemented, and an independent, sourced measurement of its real-world effect exists (e.g. a GAO or CBO follow-up report). | | `blocked` | Enacted or in progress, but formally halted (e.g. enjoined by a court, vetoed and not overridden, funding rescinded). | | `superseded` | A later law or reform materially replaced this one's mechanism. | | `unverifiable` | A real-world status change was reported but could not be confirmed against a primary source in this environment (see the WebFetch-blocked constraint documented in every research package this milestone). | | `pending_review` | A status change is suspected but has not yet been researched. | ## 3. Computation rule (v1) **Every disposition is computed, never hand-typed.** The scorecard page computes each reform's disposition directly from that reform's own public record at page-load time, using this rule and no other: 1. If any entry in `current_legislation` has a `status` value in the enacted-signal set (`enacted`, `signed_into_law`, `became_law`), the reform's disposition is `enacted`, citing that specific bill. 2. Otherwise, the disposition is `proposal`. If `current_legislation` contains one or more real bills tracking the same reform, the furthest-advanced bill's own `status` (e.g. `introduced`, `reported`, `passed_one_chamber`) is shown as context under the `proposal` badge — this is real legislative-tracking data already collected during each reform's own research pass, not a new claim. No reform is currently computed as `implemented`, `measured_result`, `blocked`, `superseded`, `unverifiable`, or `pending_review` — not because those states are unused architecture, but because, as of 2026-08-10, **every reform on this platform is a published-or-draft proposal with no enacted implementing legislation identified.** That is the honest current state of the Plan, not a placeholder pending future content. ## 4. What would change this A reform's disposition changes only when its own `current_legislation` array (or, for the later dispositions, a dedicated evidence record — see §5) is updated through the platform's normal research pipeline: a real bill number, a real `status`, and a real source citation. There is no manual "set disposition to X" control anywhere in this system — the badge a visitor sees is always a direct, reproducible function of the reform's own public data, matching this platform's standing rule against creating fake scores merely to populate a UI. ## 5. Future extension: implemented / measured_result / blocked / superseded Those four dispositions require evidence beyond a bill's legislative status — specifically, a primary-source record of executive/agency implementation action, or of an independent measured outcome, or of a court/veto/funding block, or of a superseding law. When such evidence exists for a specific reform, it will be added as a structured record validated against `data/schemas/scorecard-entry.schema.json` (already defined, currently unused) rather than by editing the computation rule to assert an outcome this platform has not sourced. `unverifiable` and `pending_review` are reserved for cases where a real-world status change is suspected (e.g. flagged in a reform's own `editorial_notes`) but has not yet been confirmed — see §3's WebFetch-blocked-environment note, which is the most likely source of `unverifiable` entries in the near term.