How the travel signals work.

One direction, different evidence

All numerical scores run from 1 (more challenges / preparation) to 5 (more favorable conditions). These are Before You Roam’s ordinal interpretations, not scientific risk probabilities or scores published by the source organizations. No overall average is calculated. Missing information stays unscored.

Inclusion Index — method version 2

This signal now reviews three lenses: ethnic/racial harassment; antisemitic harassment or crimes; and anti-LGBTQ+ harassment, violence or identity-related detention. Use the identity selector above the table to view one lens independently. The default headline is the lowest available scored lens, never an average that could hide harm affecting one group. It is a conservative editorial flag, not a validated statistical composite or an estimated probability for a visitor.

Where country data exists, resident surveys are kept separate: FRA’s 2022 survey of residents of African descent (2023 report, Figure 23); FRA’s 2023 survey of Jewish respondents (2024 report, Figure 23); and FRA’s 2023 LGBTQ+ survey (2024 report, Figure 26). Within each lens, the share reporting at least one hate-motivated harassment incident in the previous 12 months maps to: ≤10% = 5; >10–20% = 4; >20–30% = 3; >30–40% = 2; >40% = 1. These are site-defined bands, not the source organizations’ ratings.

These surveys have different populations, questions, sampling frames and years. Online harassment may be included; they are not equivalent measures of street crime and are not directly comparable tourism samples. The antisemitism survey predates 7 October 2023. The racial survey does not represent every ethnic group, including Asian visitors. A country with more reporting or broader evidence is not necessarily more dangerous. Do not interpret the lowest-band summary as a cross-country safety league table.

Outside survey coverage, explicit official identity-related visitor warnings can support a separately labeled advisory concern band: severe criminalization / detention exposure = 1; documented targeted violence or explicit detention warnings = 2; explicit harassment warning = 3. These concern bands are not incident-frequency measurements. Legal recognition or general social attitudes alone never imply a low or high harassment score. Read the geographic scope; a law or incident elsewhere is not applied automatically to a destination.

Each profile displays all three lenses, their source dates and coverage. Unreviewed lenses stay unscored and are never treated as favorable. A single report such as Kotor’s bus incident remains a localized alert, not a country-wide prevalence estimate. Broader scores may differ from the earlier single-survey version because the evidence scope has changed.

Theft and violent crime — kept separate

We manually reviewed Canadian travel guidance and UK guidance for Canada. Wording maps to editorial precaution bands: explicit low or rare = 4; reported / occurs without a prevalence estimate = 3; common, frequent, regular or prevalent = 2; explicit high or severe violent-crime levels = 1. No destination receives 5 solely because a warning is absent. If the guidance does not characterize the relevant crime well enough, it stays unscored. Occurrence-based bands are conservative editorial judgments, not comparable incidence statistics.

A warning about one neighborhood or remote region is not automatically a city-wide fact: scope appears beside the evidence. No time trend is inferred without comparable dated observations. Florence’s theft warning does not establish rising theft or high violent crime.

Water & sanitation

World Bank indicators SH.STA.BASS.ZS and SH.H2O.BASW.ZS use WHO/UNICEF JMP estimates. We take each country’s latest non-null observation in a 2020–2025 query and use the lower of the two percentages. ≥99% = 5; ≥95% = 4; ≥85% = 3; ≥65% = 2; lower = 1. The underlying value and year are shown. These measure population access to at least basic services, not drinking-water potability, restaurant hygiene or street cleanliness. They are national proxies, not city inspections.

English accessibility

EF EPI 2025 bands map directly to 1–5: very high (≥600) = 5; high (550–599) = 4; moderate (500–549) = 3; low (450–499) = 2; very low (<450) = 1. The sample consists of volunteer adult test-takers and is not representative of all residents or tourist workers. English proficiency is not foreigner friendliness or a measure of other languages. Countries absent from the reviewed index stay unscored, including primary English-speaking countries.

Credit-card acceptance

This index looks at how easily a visitor can use Visa, Mastercard, American Express and Discover credit cards directly in everyday local shops. More broadly accepted networks and fewer payment steps support a higher score.

5 = broad direct use across all four networks; 4 = credit cards broadly usable with some network or merchant gaps; 3 = mixed card / cash use or app setup commonly needed, such as Alipay or WeChat Pay in mainland China; 2 = cards usable in limited places while cash is usually needed; 1 = cash / local currency only or foreign credit cards effectively unusable.

Sources often describe general coverage rather than a census of local shops. These are editorial bands, not percentages of all merchants. Each profile shows what is known about individual networks; unverified networks are not assumed accepted. Debit-only payments do not increase this index. Missing general evidence remains unscored.

Dates, confidence and community input

Research was retrieved and reviewed on 19 October 2026. Source publication, survey and observation years differ and are shown in the details. This is a manually researched snapshot with no automatic feed. Confidence describes evidence or traveler relevance; it is not statistical certainty. Community reports are not yet a verified public dataset. You can email corrections and source suggestions to info@beforeyouroam.com. Reports need review, dates and geographic context before they inform a score.