# Methodology — Trans Travel Safety Map Each score answers one question: **how much risk does a transgender visitor face in this jurisdiction if they are, or are discovered to be, trans?** Scores run from **0.00 to 1.00, higher = safer**. A score is a comparative judgment built from researched evidence: a rough placement of one jurisdiction against another. It is not a measured probability and not a prediction for any individual trip. ## The scale - **Exactly 0.00** means a discovered trans visitor faces near-certain imprisonment **and** likely execution — both, not either. - **No other point on the scale has a defined precise consequence.** A score of 0.55 corresponds to no specified outcome, and a small gap between two scores is not a measured difference in what would happen to anyone. - **1.00** is the top of the index. No place is claimed to have reached full legal and social acceptance. - **The colour bands** (Severe Risk, High Risk, Elevated Risk, Reduced Risk, Low Risk) are display groupings for reading the gradient, not calibrated thresholds. Two places inside one band can differ materially in what their sources describe. ## The reference traveller The reference traveller holds a passport from their home country, visits shortly rather than residing, is an adult unless a claim states otherwise, and arrives through ordinary points of entry. Where a rule depends on age, care needs, document type or entry route, the record says so rather than assuming the most or least favourable case. ## Conditional on discovery — no credit for stealth The score rates what happens to a visitor who is identified as trans, not the odds of being identified. Passing, silence and avoidance earn no credit: they belong to the traveller and the itinerary, and a document check, a record query or an emergency can defeat them. Where a jurisdiction engineers discovery itself — mandatory marker disclosure, registration, surveillance, checkpoint checking — that is exposure in its own right. ## What is scored The A–K taxonomy (full version in [`research/outing-risk-taxonomy.md`](research/outing-risk-taxonomy.md)): - **A** Border, transit & security screening (document mismatches, scanners, device searches) - **B** Documents & everyday bureaucracy (hotel check-in, SIM registration, ID spot-checks) - **C** Gendered spaces (bathrooms, changing rooms, saunas, women-only transport, **hospital-ward placement**) - **D** Public presence & social reaction (harassment, filming/doxxing, service refusals) - **E** Violence & predation (hate violence, dating-app ambushes, blackmail — travellers are soft targets) - **F** Police interaction (stops, harassment, entrapment, extortion; protector vs. perpetrator) - **G** Law & criminal exposure if outed (expression/impersonation statutes, morality laws, propaganda laws) - **H** Arrest, detention & prison placement (misgendered facilities, "verification" exams, deportation) - **I** Health & medication (HRT import/customs, emergency care, insurance) - **J** Family & diaspora exposure (visiting-relatives scenarios, forced "conversion") - **K** State & media climate (a modifier of the picture above, never a standalone driver) **Not scored**: general LGB rights per se; resident-only trans issues (a visitor uses their home passport and never enters the destination's gender-marker process); general crime, terrorism, disease or conflict, except where they independently amplify trans-specific risk. The map rates trans-specific visitor risk, not general travel safety — general risks still matter in practice and appear as context in popups where sources raise them. **The visitor/resident boundary cuts both ways.** The absence of resident-facing machinery — no legal-gender-recognition regime, no anti-discrimination statute — is a resident gap, not a visitor exposure. Where it leaves visitor questions open (which documents are believed at entry, how detention would place someone), an open question is weighed as an open question: much lighter than a documented traveller-facing exposure, and never proof of safety or danger. A protection is credited only when the record shows it in force; an advocacy submission's stated aims are not law. Institutional failure is exposure evidence, not an offset. ## How evidence is weighed - **Research-driven.** Jurisdictions are assessed from collected sources, not model recall. Popups list sources where provided; a record may carry per-source summaries, a bare link, or little local material at all. - **Law, enforcement, actor, date and scope are read separately.** A bill is not law. A statute without documented enforcement weighs differently from documented arrests or police practice. Harm is attributed to whoever the record shows responsible — a party rally, a personal video or supporters' threats are not government policy unless the record makes them one. Facts carry their date. Every claim carries its scope: who it applies to, under what trigger, in which places. - **Individual incidents count for what they prove, not how shocking they are.** A single report barely moves a large country; patterns, state implication, impunity or victim-proofing, and very small jurisdictions change that. A good institutional response — convictions, public outcry, reform — counts as positive evidence. - **Burden-of-proof symmetry.** A documented positive counts as much as a documented negative. Absence of evidence is not evidence of absence — in either direction. Where sources run out or disagree, the uncertainty stays in the record instead of being resolved toward safety or danger. - **Curiosity is not hostility; hostility is not state threat-framing.** Staring, misgendering and conservative disapproval are friction: the traveller is an oddity, not a threat. Hostility frames the traveller as a deceiver or a danger to children or moral law. When law and institutions themselves treat trans people as a threat, that state framing follows the traveller and weighs heaviest. - **Sexuality law is not trans law.** A same-sex-conduct statute is secondary for trans visitors: it matters through how it is used against visibly gender-nonconforming people. Trans-specific law and practice dominate. - **A violence rate needs a denominator.** Killings at or below the general homicide rate are background violence; an admitted undercount says nothing about magnitude. Both are omitted rather than given false precision. - **Severity dominates probability.** Everyone carries some outing probability; what separates places is mostly what happens after discovery, and whether the state is perpetrator, bystander or protector. ## Country and regional assessments A country's score is its own country assessment. It is **never** derived from regional scores; a national-to-regional spread is a diagnostic observation, not a target for adjusting anything. A regional assessment begins with the **full national picture** and then considers what is locally different — added restrictions or protections — with local facts weighed by the same rules above (scope, date, trigger, visitor applicability; local evidence never doubles as national evidence). Where no local signal exists, the region carries the national score and its popup says so. That is an evidence gap, not a finding that conditions there match the country. Sub-national variation can be large: a national score cannot describe every local difference, and decimal differences and ranks should not be read as precise measurements of relative safety. ## Boundaries and sources Boundaries are **de jure** — internationally recognised legal claims, not lines of current control (Crimea and occupied territories render within Ukraine; Kashmir along the Line of Control). Patches are documented in [`tools/patch_dejure.md`](tools/patch_dejure.md). Country and first-level boundaries: [geoBoundaries](https://www.geoboundaries.org/) gbOpen (CC BY 4.0); Western Sahara polygon from Natural Earth (public domain). Research notes live in [`research/`](research/). ## Using the map This map is **informational research, not personalised travel advice**. Risk depends on the individual — presentation, documentation, itinerary, companions — and conditions change; verify against current official travel advisories before travelling. The map, its scores and project data are dedicated to the **public domain (CC0 1.0)**; boundary data retain their upstream licences.