# Responsible America — Citizen Action Policy Status: policy document, phase 1 (design intent). The Citizen Action Center itself is **not built yet** — no code exists under `responsible-america/` for it, and no representative-lookup or contact-routing data exists under `data/` as of 2026-07-30 (Phase 4 of `docs/roadmap.md`). This document defines what it must do, and must not do, before a single line of it is written, so that implementation has a fixed target rather than being improvised feature-by-feature. ## 1. Purpose Reading about a problem changes nothing by itself. The Citizen Action Center exists to close the gap between "a user understood a reform on this platform" and "a user lawfully contacted the specific person who has authority to act on it." It is not a petition tool, not a donation tool, and not a place to vent — it is a directory and a set of lawful, respectful mechanics pointed at the right office. Every reform record's `citizen_actions` and `contact_targets` fields (`data/schemas/reform.schema.json`) exist because of this purpose: a reform page that only explains a problem, without telling the reader who to contact about it, has done half the job the platform exists to do. ## 2. What the Citizen Action Center will help users do Once built, the Center is intended to help a user: - Identify their federal representative and senators, their state legislators, their governor, and the relevant congressional or state committee for a given issue. - Identify the specific agency official or office responsible for administering a given policy area. - Identify their local representatives (city council, county commission, school board, etc.) where a reform is locally actionable. - Locate official contact forms, published office phone numbers, publicly listed email addresses, and district office locations for the above. - Understand which official actually has jurisdiction over a given reform, so effort isn't wasted contacting an office with no authority over the issue. - Send a respectful, substantive message to the right office. - Submit a formal public comment during an agency's public comment period. - Testify, or submit written testimony, at a legislative hearing. - Contact committee staff (not just the elected member) when staff is the more effective point of contact for a given ask. - Find and join relevant nonpartisan organizations already working on an issue, as a way to act collectively rather than only individually. - Track a specific bill's status through the legislative process. - Monitor how their representatives voted on relevant legislation. - Find and attend public meetings (city council sessions, town halls, agency listening sessions) relevant to a reform. - Request public records (e.g., FOIA or state equivalents) relevant to a reform or an official's conduct. - Report suspected misconduct through the correct lawful channel (an inspector general, an ethics office, a licensing board) rather than through public accusation on the platform itself. None of this exists yet. This list is the design target for Phase 4, not a description of current functionality. ## 3. `contact_targets`: one specific, actionable target per reform Every reform record should specify, via `contact_targets` (`data/schemas/reform.schema.json`), the single most effective advocacy target for that reform — the one office or body a reader should contact first if they only take one action. A good `contact_targets` entry is specific enough that a reader (or the eventual Citizen Action Center lookup tool) can act on it without further research. Compare: - Good: `"your U.S. Representative's office"` — a defined, findable role, even though the specific person depends on the reader's district. - Good: `"the Federal Election Commission's public comment docket"` — a named body and a named mechanism. - Good: `"the House Judiciary Committee's Subcommittee on the Constitution"` — a specific committee with jurisdiction over the issue. - Good: `"your state's Secretary of State (election administration)"` — a specific office defined by function, since the exact office name varies by state. - Bad: `"the government"` — not a target, not actionable. - Bad: `"Congress"` — too broad; doesn't tell a reader which chamber, committee, or member has jurisdiction over this specific reform. - Bad: `"your elected officials"` — doesn't narrow to the office that actually has authority over the reform's subject matter. The test for a `contact_targets` entry: could a reader, or a lookup tool, resolve this to an actual mailing address, phone number, or web form without having to first figure out who has jurisdiction? If not, it needs to be narrowed before the record can pass `research_review`. ## 4. Hard prohibitions These are bright lines, not judgment calls to be balanced against other goals. The Citizen Action Center — and any tooling, template, or generated text associated with it — must never generate, suggest, or facilitate, even indirectly: - **Harassment** of any official, staff member, or private individual. - **Threats** of any kind, explicit or implied. - **Doxxing** — publishing or aggregating an official's or private individual's personal information (home address, personal phone number, family members' identities, etc.) beyond what that person has made publicly and officially available in their capacity as an official. - **Mass spamming** of a single office — including auto-generating a form-letter or mass-blast message template designed to overwhelm one office's inbox or phone lines with near-identical messages. A tool that helps many users each write their own genuine message to the correct office is in scope; a tool whose actual function is to flood one office regardless of message substance is not, and must not be built even as a "convenience" feature. - **Impersonation** — of an official, of another citizen, or of an organization. - **Unlawful interference** with government proceedings, offices, or personnel. - **Contacting private family members** of officials. Contact is limited to the official in their official capacity, through official channels. These prohibitions apply to the tool's own behavior (what it generates or automates), not only to what a user might do with it on their own initiative. If a future feature's most natural implementation would cross one of these lines — for example, a "one-click send to every representative in a district" button — the feature must be redesigned or dropped, not shipped with a disclaimer. ## 5. Public comment / deliberation mechanism (forward-looking) Not built yet. When it is, it will not be an open, unmoderated comment section. Instead, public engagement with a reform will use **structured reactions**: - Support - Oppose - Support with changes - Submit objection - Submit alternative - Identify unintended consequence - Suggest source - Report outdated information - Propose compromise This structure is deliberate: it gives the platform real signal about where public reaction actually lands (agreement, disagreement, a specific factual objection, a proposed compromise) without opening a free-text space that becomes the default venue for harassment, off-topic argument, or spam. Free-text input, where it exists at all inside these structured categories (e.g., explaining *why* someone opposes a reform, or describing an unintended consequence), is scoped to that category's purpose, not a general comment box. **Public submissions are always kept visibly separate from approved Responsible America positions.** A reform page's own content (problem statement, proposed solution, supporting/opposing arguments as drafted and reviewed per `docs/governance.md`) is the platform's editorial position. Structured public reactions are displayed as public input *about* that position, never merged into or presented as part of the position itself. A reader must never be able to mistake "100 users selected 'oppose'" for a Responsible America editorial judgment, or vice versa. ### Moderation Moderation of structured submissions prohibits: - Threats - Targeted harassment - Personally identifiable information (PII) about any person - Discriminatory abuse - Spam - Fabricated evidence - Unlawful content **Viewpoint disagreement alone is never grounds for removal.** A structured "oppose" reaction, a well-formed objection, or a factual counter-source is not moderatable content merely because it disagrees with the platform's editorial position — moderation acts on the prohibited categories above, not on the position taken. ## 6. Correction-request path An ordinary reader who spots an error, an outdated statistic, or a broken source link on a reform page needs a way to flag it that does not require understanding the platform's internal review workflow. That flag is the "report outdated information" (or equivalent factual-correction) structured reaction described in Section 5. What happens after a correction request is submitted — how it enters the claim-verification queue, what verification states a flagged claim can move through, and who is responsible for resolving it — is defined in `docs/fact-checking-policy.md`, not here. This document only establishes that the entry point for an ordinary reader is a simple, one-click structured flag, not a requirement to file a github issue, write a formal letter, or navigate the editorial workflow states described in `docs/governance.md`. Once `docs/fact-checking-policy.md` exists, it is the authoritative source on what happens after that flag is submitted; this document should not be read as defining or duplicating those verification states. ## 7. Relationship to reform records - `citizen_actions` (`data/schemas/reform.schema.json`): a list of concrete actions a reader can take related to this specific reform, drawn from the categories in Section 2 (e.g., "submit a comment to the FEC's rulemaking docket on X," not a generic restatement of Section 2's list). - `contact_targets`: the specific advocacy target(s) for this reform, per the standard in Section 3. Both fields should be populated during `research_review` / `editorial_review` (per `docs/governance.md`'s workflow states) as part of making a reform record complete, not left as an afterthought once a record reaches `approved`.