GayOut Index — Methodology

How we score countries on legal safety and the LGBTQ+ scene that actually operates there.

The GayOut Index combines two things usually measured separately: what the law says, and what exists on the ground. Legal indexes tell you where you can be arrested. They don't tell you whether there's anywhere to go on a Saturday night. We publish both, together, because a traveler needs both.

GayOut Index = 60% Legal Safety Score + 40% Scene Score.

We weight law above scene deliberately: criminalization outranks nightlife. A country where being gay is illegal is not made safe by having bars. The 60/40 split is a judgment call, not a derivation — the sensitivity table below shows how the rankings move under different weightings, so you can judge how much the choice matters.

1. Legal Safety Score (60%)

Scored 0–100. As of release 2026.3 this is GayOut's own composite, built directly from primary sources — no other travel index enters the formula. The legacy industry index conversion is retained in the dataset as a benchmark column (legacy_benchmark_0_100) for comparison only. Gender-recognition and enforcement components were researched per country (Aug 2026) against ILGA World, Human Dignity Trust, Equaldex, Rainbow Europe and 2024-25 human-rights reporting. Components:

The enforcement component exists because law on paper and law in practice diverge in both directions. Where enforcement evidence contradicts the statute book, the enforcement component moves, the statute components don't — the disagreement stays visible instead of being averaged away.

2. Scene Score (40%)

Computed from the GayOut directory — 17,000+ venues across 950+ cities — along three dimensions that measure different things:

Why these three

Earlier drafts blended per-capita density with absolute venue counts. Those two correlate almost perfectly, so the blend silently collapsed into "who has more venues." Density, spread and diversity are genuinely independent: a country can be dense but concentrated (Malta), spread but thin (Brazil), or large but uniform. The three-way split keeps each failure mode visible.

Computation

• Urban population (UN World Urbanization Prospects), not national — scenes are urban; a national denominator punishes large rural countries for having countryside.
• Log scaling because venue counts are heavy-tailed: the difference between 5 and 50 venues matters more than between 500 and 545.
• No free ride without a scene: zero listed venues = zero scene points — a country cannot rank high on law alone. Below the 10-venue floor the scene component is a provisional partial credit (0–25, scaled by venue count) and the row is marked thin-scene-provisional.
• City-states (fewer than 5 listed cities): the spread dimension would punish geography rather than the scene, so it is dropped and the remaining dimensions are renormalized (46.7% density / 26.7% diversity / 26.7% culture).
• Underground-scene discount: where the enforcement component is active or severe, the scene contribution is capped at 10 of its 40 points — venues operating under persecution are survival infrastructure, not a travel asset, and must not lift a hostile country's ranking.
• Gay bars are counted only when LGBTQ+-dedicated; "gay-friendly" listings are excluded from all scene metrics.

3. Coverage — read this before comparing countries

The Scene Score measures what is in our directory, and our directory is not equally complete everywhere — it is deepest in Europe and the Americas. A per-country Coverage Grade (A–D) (share of venues verified in the last 12 months, number of independent source types, date of last full editorial review) is published in the dataset now (grade = 0.6×share of venues verified in the last 12 months + 0.4×share with a verified place-ID). Coverage is displayed next to the score and is never folded into it — folding it in would reward us for our own thoroughness rather than the country for its scene.

4. Update cadence & versioning

• Quarterly frozen releases — 2026.3 (Sep), 2026.4 (Dec), archived at /downloads/gayout-gti-2026-3.csv etc. The frozen file is the citable file: it reads the same when your editor checks it next month.
• Live directory figures (venue counts, events) refresh weekly and are labeled current.
• Legal changes between releases are annotated on the index immediately and folded into the next frozen release.

Changelog

5. Sensitivity — how much do the weights matter?

The 60/40 split is a choice. What happens to the top-20 ranking under the alternatives (computed on the current live data, refreshed weekly):

If small weight changes reshuffled the whole board, the composite would be telling you more about our arithmetic than about the world. Publishing this table lets you check.

6. Known limitations

• Coverage bias. Deepest in Europe and the Americas; the Coverage Grade will expose this, not eliminate it.
• Venue-type skew. Directory categories historically over-represent venues aimed at gay men. Scenes centered on women, trans and non-binary communities are under-counted; treat the Scene Score as weakest on this axis.
• A listing is not an endorsement of safety. The Legal Safety Score, not the venue count, is the safety signal.
• Sub-national variation. A single national number hides São Paulo vs the interior, or differences between US states. City pages carry the local picture.
• Timing. Legal components are reviewed quarterly; a law passed yesterday may not be reflected until the annotation appears.

7. Cite this index

Research & methodology: Ziv Goldvasser, M.B.A. (Tel Aviv University) — founder, GayOut — with the GayOut editorial team (publishing LGBTQ+ travel data since 2008) and quarterly legal review against the sources listed above. Found an error? Write to info@gayout.com — corrections are logged in the changelog.

← Back to the Gay Travel Index